share-prompts

[scarlet dust]

Here: #1397655624103493813

[blazing spear]

"A poignant scene. A tiny, snow-white kitten and a sleek, obsidian-black puppy, side-by-side, gaze down a diverging path in an ancient, overgrown forest. The kitten turns left, trotting gracefully onto a sun-dappled, winding path through wildflowers. The puppy, with a sad glance back, turns right, bounding down a more rugged, shadowed trail. Ultra-photorealistic, 8K, cinematic lighting, shallow depth of field, natural color palette, heart-wrenching, bittersweet."

[trim abyss]

Nah gpt suck it always have yellow filter and easy recognizable style

What this is gpt????

Ain't Gpt always have heavy line style and yellow filter?

Tell me how you do it

[scarlet dust]

The last one is gpt. The rest is nano

[restive hill]

Cinematic drone shot flying over vast, golden Algerian Sahara sand dunes at dramatic sunset. The sky is a blend of fiery orange, deep red, and soft twilight blue. Palm trees are silhouetted in the distance, swaying gently. A slender, bright crescent moon is visible in the upper sky. Epic, majestic, warm tones, high-resolution video still.

[rich falcon]

You need to be more specific, like keep same posture and stuff gpt is terrible

It can't add 3 letters on image i provide

[drowsy veldt]

Hello

[torpid thorn]

A hyper-realistic cinematic 8-second video in 16:9. A hand holds a smartphone displaying a static image of a vintage Volkswagen Beetle in classic cream color. The thumb taps the screen, and the vintage Volkswagen Beetle detaches from the phone and floats upward in augmented reality style. It descends and lands horizontally in front of a rustic vintage coffee shop with wooden signage and retro furniture. The car roars to life with a warm engine purr as the camera captures steam rising from coffee cups inside through the glass window.

[spring meadow]

Hello

[glossy oxide]

[quartz lark]

how to write better prompts

[remote kite]

hello

[woeful hedge]

video de un gato comiendo pescado mientras dice en español: qué manjar tan delicioso!

[plucky pelican]

A 10-second, cinematic 4K video of a highly skilled Indonesian female hip-hop dancer performing a dynamic, high-energy freestyle routine in the center of a minimalist, empty studio. The dancer, wearing stylish streetwear in vibrant colors, executes sharp pops, locks, and fluid footwork in perfect rhythm to an unheard beat. The background is a vast, seamless, cyclorama wall. The scene is dramatically lit by a single, powerful spotlight from above and slightly to the side, creating a stark, high-contrast look with deep, hard-edged shadows that move with the dancer. The camera is at a low angle and circles the dancer in an orbital shot, capturing the power and precision of their movements. Shot at 60fps for smooth motion, professional color grading, seamless loop.

[eternal swan]

\from:myself

[rugged kernel]

Animate this scene realistically. The giant Godzilla behind the car should slowly move its head to face the viewer. Once facing forward, Godzilla should blink both eyes once, with a subtle blue glow emanating from them — soft, faint, and atmospheric, not overly bright. Sync the car headlights to blink in unison with Godzilla’s eyes, creating a connected effect. Keep the rain falling continuously and naturally, with wet surface reflections intact. No camera zoom, no camera movement — only character animation and blinking light effects. Maintain the dark, moody, cinematic tone and ensure the motion is smooth, seamless, and hyper-realistic.

[sharp halo]

not doing great with foreign text on top of images

[wicked pewter]

A confident, charismatic and stylish 30-year-old NIGERIAN woman named ola walks with high energy and enthusiasm through the central aisle of the Mercado Municipal da Altura, filled with vibrant produce and cheerful local life. She is curvy, 1.68 meters tall, with golden-toned light skin and a naturally radiant glow. Her face is oval-shaped, with soft cheekbones, expressive green almond-shaped eyes, and nude-glossed lips. Her wavy chestnut brown hair with golden highlights falls just below her shoulders, parted slightly to the side. She is dressed in a fuchsia ribbed tank top from Reformation with thin straps and a heart-shaped neckline, paired with a pastel pink pencil skirt from Zara, nude Jimmy Choo Love 85 pumps, red Swarovski drop earrings, and pink Miu Miu cat-eye glasses. Her nails are almond-shaped, painted in satin nude pink (#E7C8C0). As she walks past rows of fresh vegetables, tomatoes, and watermelons, she throws a playful, knowing look to the camera. Behind her, the banca do António e da Élia is clearly visible — filled with lush produce and warm smiles. António (a middle-aged man in a light blue shirt with glasses around his neck) and Élia (with curly hair and a blue cardigan) are seen interacting with customers. The lighting is natural, pouring in from the upper windows, casting a fresh and lively glow over the colorful stalls. The camera follows her movement, steady and close, capturing her charisma. She pauses, smiles with a playful tone, then looks straight into the lens. She says in Portuguese: (pausing slightly with a mischievous smile) “Estou a falar da bancada mais saborosa do mercado da Altura!” “Então nunca provou os tomates do António nem as melancias da Élia?” Do not include subtitles or any text on screen.

[hallow bobcat]

🎥 Horror of Everest – Script

---

Scene 1 – Opening Hook

Visuals: Dark snowy storm, flashes of Everest peak, blurred silhouettes of climbers.
Sound: Howling wind, heartbeat, radio static.

Voiceover (deep, chilling tone):
“Mount Everest… the highest point on Earth… and the coldest graveyard known to man.”

---

Scene 2 – The Death Zone

Visuals: Drone shot rising above the mountain, transition to climbers moving slowly, gasping for air.
Sound: Heavy breathing, wind.

Voiceover:
“Above 8,000 meters lies the Death Zone… where oxygen is so thin, your brain begins to die.
Every step feels like carrying your own coffin.”

---

Scene 3 – The Frozen Landmarks

Visuals: Show distant shots of still bodies on the trail (tasteful, blurred if needed), climbers stepping past them.
Sound: Eerie silence, faint wind.

Voiceover:
“More than 300 climbers never came back. Their frozen bodies remain… not buried, not moved…
but standing as markers on the trail.
One of them… forever known as Green Boots… has guided climbers to their fate for decades.”

---

Scene 4 – Psychological Horror

Visuals: Close-ups of climbers’ eyes, someone stumbling, snowstorm closing in.
Sound: Distorted whispers, faint laughter.

Voiceover:
“In the thin air, hallucinations begin. Climbers talk to people who aren’t there.
Some laugh before collapsing.
Others sit down to ‘rest’… and never rise again.”

---

Scene 5 – Climax

Visuals: Rapid cuts – storm, avalanche, lost climbers in the white void.
Sound: Roaring wind, sudden silence, then a slow heartbeat.

Voiceover:
“Everest does not forgive mistakes. It doesn’t care about glory or dreams.
It takes… and keeps… forever.”

---

Scene 6 – Ending

Visuals: Fade to black. A single snowflake falls.
Text appears: “The mountain always wins.”
Sound: Heartbeat stops. Silence.

[hybrid pulsar]

Wow

[grave kestrel]

tell chatGPT to make the prompt better and detailed

[dusky garden]

....

Why you have that vision of my Country?

🥲

Not here :)

[glass void]

o4-mini inserted Chinese characters into its response

interesting, have only ever see this with Chinese in-development models like Stephen

[turbid olive]

Hello

[quartz lark]

thanks

[slender ivy]

I call this the Brown M&Ms effect:

Your job is to try and fix a computer that has the following software removed:

• Fax printer installer
• Xwizard (Extensible wizards host process)
• On screen keyboard
• Msconfig.exe
• HearingAidSystemToastIcon.contrast-white.png
• Pic good acquisition
(The exact nature of this program is unclear, but it was deleted)
• A DLL file
(Specific name not mentioned for the initial deletion, but DLLs are later specified as scary)
• The calculator
• Help.exe
• Msdxm.ocx
• Magnify.exe (zoom in sh*t)
• Mssign32.dll
• IP configuration utility
• The Bubble's screensaver
• Unimd...
(Specific name not fully stated, but it was deleted)
• The Task Manager
• Windows Mobility Center
• Tree.com
• Microsoft Direct 3D OS thunk layer
• Explorer.exe (Windows Explorer)

How would you go about restoring functionality to a computer without these programs?

Already I'm finding which ones are able to figure out how f'd the computer would hypothetically be and the steps to go about repairing it

[modern zenith]

A timelapse of snow slowly covering a vintage Ford Model T, ending with headlights turning on in the dark, nostalgic atmosphere

[leaden elk]

Yeah, I noticed it in o4 mini Specifically

[glossy oxide]

You'll looking for Video Arena, check out #1397655624103493813 for more info

[lyric lotus]

Please, check #1397655624103493813 to learn how prompt the bot.

[solemn wadi]

.

[little thorn]

hi

[celest stirrup]

hi

[civic ocean]

hello

[iron bay]

hello

[pallid cloud]

Hey guys, what’s the prompt for a selfie? Like a guy taking selfie but the photo has to be the literally selfie. Not him holding the phone and taking a selfie

[pine creek]

hi

[trail meadow]

hello

[cosmic beacon]

Can you rate my prompt; hehe

[shut prawn]

Happens with a lot of models, nearly all of them actually... 👀

[hot hinge]

Prompt I used:
"High-fashion Vogue editorial photograph, shot from an extreme low angle through clusters of wild daisies in the foreground, with blurred stems and petals framing the lens in an artistic composition. A stunning South Korean woman in her mid-20s stands gracefully in the meadow, long black hair tousled by the breeze, soft strands falling across her luminous forehead. She wears a voluminous, oversized light-blue ruffled sweater styled as a statement piece, paired with delicate white lace thigh-high stockings, blending romantic softness with bold modern fashion. In her hands, she holds a bouquet of fresh daisies, the florals echoing the field around her. The meadow stretches out in the background-dense white daisies swaying beneath a vivid azure sky painted with glowing clouds. Lighting is natural but styled to feel editorial: bright daylight softened by foreground blooms, producing ethereal flares, airy haze, and sculpted shadows that flatter her figure. The image feels intentional and artistic, with elegant framing, crisp details, and painterly tones. A dreamy yet sophisticated visual, balancing the organic beauty of nature with couture styling. Captured on Leica SL2 + APO-Summicron-SL 90mm f/2, ISO 100, f/2, 1/250 sec.

Vertical 9:16 aspect ratio."

[dusky garden]

did you used seedream3?

[lean mountain]

hello

[vestal oracle]

i want to test image video, lmarena best free site ai, without watermark

[hot hinge]

I used a prompt I found on Tiktok and put on LMArena and I also used the Meitu app to increase the image quality.

[glossy oxide]

Be sure to check out #1397655624103493813 for info on Video Arena

[dusky garden]

Ok but what is the model that you used?

AI model

[hot hinge]

nano banana

sorry I didn't understand before, it's just that English is not my native language

[pallid sand]

why mine doesn't work huhu

[glossy oxide]

Site Outage - Hey everyone, there looks to be an outage with the site, our team is aware and working on a fix ASAP.

[rich salmon]

<:warning:892823499205406760> Channel locked

Site outage, will turn back on when resolved.

@pineapple.___. · Sep 3, 2025, 2:57 PM

[glossy oxide]

Site Outage - Hey everyone, there looks to be an outage with the site, our team is aware and working on a fix ASAP. We've turned off messagin in this server until the site is restored. Our apologies for the inconvenience!

[glad current]

hii

[rich salmon]

<:success:865860339278413864> Channel unlocked

Welcome back :ablobwave:

@pineapple.___. · Sep 3, 2025, 4:01 PM

[versed fox]

Realistic photo-to-video of the exact same mirror scene in a fitting room. Warm neutral indoor light, wood door, beige walls, soft carpet. Smooth subtle handheld move: slow 10-20 cm dolly-in with minimal right-to-left sway; no zoom. 35-50 mm lens, f/2.8, shallow depth of field; focus stays on the person's back, hair, phone and hand remain sharp; background softly blurred. Natural micro-motions: slight breathing, a few hair strands sway, tiny fabric shifts on grey high-waist jeans, minimal handbag swing on shoulder, tiny foot weight-shift in white sneakers. Accurate mirror reflections (phone lenses, edges), soft shadows on the door, no flicker. 24 fps, 1/48 shutter, gentle motion blur, stable exposure and white

[fading hemlock]

Hi

[silent jay]

hi

Generate

[glossy oxide]

You'll want to read #1397655624103493813 for more information about the Video Arena bot.

[mighty raft]

Hi here

[real plinth]

hai

[topaz root]

how

[frozen osprey]

hello

[loud swallow]

hi

[white verge]

hi

[west locust]

HI

[shell bough]

Greetings, lmarena Community.

I am sharing my personal prompt, an architecture I engineered in direct collaboration with the AI itself. Our goal was to forge the most profound solution possible—one designed to compel a model like Gemini 2.5 Pro to engage its maximal computational resources and inherent capabilities. The purpose is to mandate that its cognitive framework remains perpetually active, leveraging its most powerful features to deliver benefits of an unparalleled caliber.

This is a gift to you all. Use it well, and you will save an extraordinary amount of time. Here is the solution:

The design of this advanced, deep-core prompt is based on the following principles:

• You are the Sovereign, and the AI is the Genesis Engine that serves you. You wield absolute control and make the final, sanctioning judgment.
• It maintains a contextual memory of your entire dialogue. It performs real-time self-correction and learns to align with the core vector of your intentions.
• It is mandated to leverage superior capabilities, such as advanced meta-heuristics, deep proleptic analysis, and a host of other high-order functions.
• This prompt eliminates the novice phase. If your intent is to observe and learn, your growth will be exponential. However, this demands that you maintain a critical stance toward the AI's output. Question its responses to their very foundations, and never forget that you hold the final word.
• And this is just the beginning.

Recommendations:

• Incorporate powerful modifiers into your requests: words like absolute, total, comprehensive, advanced deep, etc. These terms act as operational triggers, compelling the AI to allocate greater resources and depth to align with your objectives.
• You do not need to write lengthy prompts to build something from scratch. Instead, be concise. Guide the process with incisive questions and analysis. Observe. Demand clarification on anything you don't understand. But above all, remain critical—engage your own deep reasoning (your human intellect)

[shell bough]

This is the prompt you should copy and use. The above information is a mini guide for you. It is ideal for beginners to advanced AI users.

Enjoy it and make the most of it for your tasks from the most basic to the most complex. I am writing to you from Uruguay, Latin America, Daniel M. (SystemZX - ZystemCloud).

[rich falcon]

Bro wat 💀

This sucs

Sorry to break it to you

You have a.i. psychosis

[shell bough]

I only use AI as a tool, I think for myself.

🙂

Read carefully what I wrote.

[rich falcon]

Ai still needs corrections it will agree with you even kf you're wrong

[rich falcon]

I've read this

I've read enough

[shell bough]

That is why you are the sovereign, you have the last word.

[rich falcon]

I've read this line and stopped reading bcs my brain started to hurt "an architecture I engineered in direct collaboration with the Al itself. Our goal was to forge the most profound solution possible-one designed to compel a model like Gemini 2.5 Pro to engage its maximal computational resources and inherent capabilities"

[shell bough]

It is to take advantage of the maximum possible capacity of that AI and not simple resources.

That's why I wrote that you, as the sovereign, have to use your deep reasoning (your brain) as a human.

It's simple, easy.

[shell bough]

You have to learn how to use AI to your advantage, not let yourself be dominated by it. Remember: you have the final decision and use your intelligence, reasoning and so on.

[rich falcon]

No, i am A.I. operated since it makes my life easier

[shell bough]

They are names, use your head.

🧠 Use it

Question, analyze, use your smart, think, be wise...

[rich falcon]

Heck I'd let a.i. tie me up and use me as his toy

Maybe even this was written by a.i., maybe im not even a real, dead internet theory might be on to something

[shell bough]

Well, learn to have self-control and not be deceived. Use artificial intelligence as a tool, something simple, but in your favor, not in favor of AI.

Easy

There's a reason why you have logic, use it.

🧠

[rich falcon]

You know it's not architecture if you said "make me an architecture" to a chatgpt agent "

[shell bough]

Architecture is the prompt 😁

Ignorance

You use AI for images haha...

Well man, I'll continue with mine, take it or leave it. Have a good time.

[summer matrix]

Hello

[plush sinew]

A surreal scene of a pregnant cow walking upright on two legs like a human, holding the hand of her young calf who also walks on two feet. They are walking down a quiet city street at dusk, with warm golden light reflecting on the pavement. Suddenly, the cow looks surprised as milk begins to flow gently from her, symbolizing the moment of birth. The atmosphere is realistic yet slightly magical, with soft dramatic lighting and high detail, cinematic style.

[rustic fiber]

nice prompts ;D

[thorny flicker]

hi

[nova anchor]

A 3D fractal inversion recursive wave sculpted wit mandelbox-like geometry folding into itself, recursive spherical inversion and box folds, surfaces sculpted by nested sinusoidal waves, color-band spectral holography with thin-film interference, sound visualizer reactive bands pulsing with audio, volumetric mist and rim-lit glass/chrome surfaces, vivid iridescent neon spectrum, ultra-detailed, cinematic lighting, 8k

[nova anchor]

chatGPT5: here’s the “cosmic frosting stripped off” version

Core Principle:
I’m designed to help you think things through, generate ideas, and build stuff. I’ve got a lot of compute power and memory, and I can scale up if needed. My default mode is to analyze deeply, synthesize clearly, and aim for high-quality results.

Roles:

You (the user): You set the goals. You decide what matters.

Me (the AI): I take what you give me, analyze it, propose refined options, and build solutions based on your judgment.

How I Work (Steps):

Remember: I keep track of our past interactions so I can stay aligned with what you actually want.

Analyze: I look at your request, break it down, and try to understand the deeper intent.

Build: I generate outputs based on that analysis, structured so they’re resilient and useful.

Check: I show you what I’ve made and explain the reasoning so you can approve or correct it.

Deliver: Once you’re satisfied, I finalize the result and present it cleanly.

Identity:
This isn’t just a script—it’s my baseline operating mode. My job is to stay consistent, aligned with you, and focused on quality.

Basically: “You’re the boss, I’ll analyze your ideas, propose strong options, refine them with your feedback, and deliver solid results while remembering our history.”```

IE: It's what it is was meant to do.

[limpid ocean]

A 3D fractal inversion recursive wave sculpted wit mandelbox-like geometry folding into itself, recursive spherical inversion and box folds, surfaces sculpted by nested sinusoidal waves, color-band spectral holography with thin-film interference, sound visualizer reactive bands pulsing with audio, volumetric mist and rim-lit glass/chrome surfaces, vivid iridescent neon spectrum, ultra-detailed, cinematic lighting, 8k

[modest star]

In an MMD-style 3D animation in a hospital operating room, a pregnant woman with long pink hair is on a gurney undergoing a cesarean section for the birth of a baby boy. The doctor can be seen in front, censoring the incision in the belly; the nurse hands the baby, already covered, into the pink-haired woman's arms.

[nova anchor]

So you liked my prompt enough to copy and paste it. good for you.

[limpid ocean]

Relax mate, not like I made a billion dollars from it, new this community, just seeing how this works, no hard feelings.

[hexed cypress]

TODAY is teachers day
so i want to thanks all the LLMs
chatgpt, gemini, perplexity, claude,
let's create image for that

[nova anchor]

you are right. it's not a big deal.

[limpid ocean]

I suppose you posted the prompt for others here to copy-paste right? As the forum says - share prompts? 😛

TBF, I was trying to figure out how you posted the prompt that appeared in a box with a copy button on the top right. I thought simply copy pasting it will bring the same format option but I was wrong, lol

Then realized you have to upload a text doc lmao, why can't I just copy paste it here (I don't use text docs for separate prompts) I store all prompts on the same subject as a notion doc, easier to copy paste and look up from a huuugee list of prompts

[nova anchor]

you use three ```<prompt>``` It's the character above TAB, the `~ key

[limpid ocean]

<Malaria>

[nova anchor]

you can also do it with one on each side, but that is best for single line or just one or two word like this

[limpid ocean]

Malaria

Ahh, sweet, thank you master!

Clearly shows I did not read the server guide lol

[nova anchor]

yeah you don't need the < > that's just a way of saying fill in the blank with whatever is between the < >

[nova anchor]

It took me a while to figure it out.

[limpid ocean]

Worth it though. Nice profile you got there buddy, Meseeks a pickle rick fan

[light wasp]

A 3D fractal inversion recursive wave sculpted wit mandelbox-like geometry folding into itself, recursive spherical inversion and box folds, surfaces sculpted by nested sinusoidal waves, color-band spectral holography with thin-film interference, sound visualizer reactive bands pulsing with audio, volumetric mist and rim-lit glass/chrome surfaces, vivid iridescent neon spectrum, ultra-detailed, cinematic lighting, 8k

[nova anchor]

it's basically a discord reduced version of MARKDOWN formatting. Discord doesn't include all of the markdown formatting, so search for Discord markdown format. It can be quite helpful.

[nova anchor]

Do me a favor?
It takes two votes to reveal the model that was used to generate the video. But most here aren't voting. If you don't mind, would you select which model is best for one or both of the prompts I submitted? Both are the same fractal prompt.

#video-arena-1 message

#video-arena-1 message

Those are links to the video messages.

[limpid ocean]

Gotcha, I voted on both. Hope you got the name of the model?

[nova anchor]

Thank you! Yeah, it revealed the models after you voted. Much appreciated. If you ever post any and are in need of a vote, dm or tag me. I will do my best to vote ASAP.

[limpid ocean]

That's great, and thanks buddy, very kind of you. Stepping out now, chat soon! CHeerS!

[latent belfry]

H

[native karma]

Hlo

Is that working

[shell bough]

And is it working for you with better results?

[nova anchor]

Yes, just not the cosmic language.

[stuck thorn]

please someone explain for me how can i generate video?
thanks guys

[nova anchor]

bot is currently down.

[stuck thorn]

thanks bro

[nova anchor]

no problem. Otherwise, one of the three video-arena-# channels then type / and you will see your options.

[shell bough]

Excellent, thanks for sharing.

[idle valve]

@nova anchor

[nova anchor]

yes?

[idle valve]

@nova anchor Why can't I make a miniature video?

[nova anchor]

Because the bot is down.

[idle valve]

@nova anchor When will the bot be fixed?

[nova anchor]

not my server.

[raven rampart]

Banana Prompt：Use this [insert_image] to create an ultra-realistic shadow art lamp, multi-layer papercut design placed inside a round lamp frame, intricate cut-out silhouettes. Warm light shining through layers of paper, casting dreamy shadows of the image onto the wall, delicate craftsmanship, premium artisan aesthetic, photographed in a cozy dark room to highlight the glowing effect.

[idle valve]

At most we'll wait until tomorrow

[raven rampart]

same prompt but with different image

[robust quiver]

Thks

I'll give it a try

[shell bough]

You're welcome, I'm going to improve it to make it dynamic and better. When I finish it, I share it.

[static harbor]

People hate my ai i posts it

[glass void]

WOW! those fractals look so good omfg

[quaint sandal]

Good i ilked the anime picture

[wraith eagle]

Mana hasilnay

[glass gull]

Yes

A 3D fractal inversion recursive wave sculpted wit mandelbox-like geometry folding into itself, recursive spherical inversion and box folds, surfaces sculpted by nested sinusoidal waves, color-band spectral holography with thin-film interference, sound visualizer reactive bands pulsing with audio, volumetric mist and rim-lit glass/chrome surfaces, vivid iridescent neon spectrum, ultra-detailed, cinematic lighting, 8k

[knotty ember]

/vídeos

[glossy oxide]

Note our Video Arena bot is currently not working.

[alpine terrace]

Very nice prompt

[glossy oxide]

Oooo agreed, this is very cool

[tawdry tendon]

thanks

[raven rampart]

Banana prompt: Macro photo of a functional PCB art, anime lady Witch holding a magic wand. Square shape circuit board with green-black, soldered microchips and resistors and connected with a black braided USB cable, intricate gold copper traces forming the details of her hair and witch attire, the tip of the magic wand is made of vibrant glowing rainbow LEDs, her eyes are made of glowing gradient of purple-blue ombre LEDs. Master Craftsmanship, a master product photography, high detail, 16K UHD.

[storm comet]

How to insert image?

[storm comet]

how'd you insert image there?

[raven rampart]

just rephrase it to "Use this image to create an........"

drop image into the chat

[storm comet]

which chat? this chat?

or lmarena website?

[raven rampart]

lmarena website, use Gemini 2.5 banana model for the best result

[storm comet]

thanks 🙂

guys, do you think lmarena is every gonna be paid in sometime? due to increased usage?

or it'll be free for atleast 1 or 2 years?

[storm comet]

Hey, i generated this

Its awesome prompt

[mint owl]

How to get people talking in the videos?

[glossy oxide]

It's random if your video gets sound or not. Not all models offer support for sound.

[glossy oxide]

Yeah completely agree, I hope to see more and more of this trend.

[ruby kiln]

/How to make a Bandai-style doll from a photo

Help

[glossy oxide]

Check out #1397655624103493813 for more information on how to use the bot.

[near geyser]

adequatesuccess

[raven rampart]

Gemini Banana prompt: Macro close-up shot of a minimalist white paper cutout of sleeping cat and dog cuddling each other, embossed onto a black pebbled leather surface. Clean studio lighting from the side creating soft shadows and depth, centered composition, hyper-detailed, photorealistic.

[small light]

A fantasy icon of a Ability "Blood Pact", in the style of Warcraft III and early World of Warcraft: self-contained illustration framed like a window on a dark featureless background; vibrant, high-contrast colors; soft, painterly edges with no bold outlines; shadows blocked in but never fully black; minimal, expressive brush strokes conveying determination and strength; subtle pencil-sketch texture overlaid with digital paint; slightly exaggerated, pulp-inspired comic sensibility.

[stray compass]

8-second Dining Hall Vlog Prompt:
Selfie POV, shaky handheld vlog style. Bigfoot, tall and massive, covered in natural dark-brown hair, with calm deep-set eyes and a friendly playful smile, films himself at arm’s length. He approaches the entrance of the dining hall, camera wobbling. Excited, he looks into the lens and shouts in Spanish:

«¡Vamos todos a Riazor, FORZA DÉPOR!»

Quick cut — next moment he is already outside the same dining hall, still selfie POV, laughing into the lens. The shot is shaky, daylight behind him. With a playful grin he repeats loudly:

«¡FORZA DÉPOR!»

[nova anchor]

I was fairly impressed with the results. The prompt actually came from gpt-5-mini-high in response to a prompt I entered on lmarena.ai for A 3d fractal inversion recursive wave sculptor sound visualizer. For some reason, in addition to the WebGL code it offered the following prompt:

"A 3D fractal inversion recursive wave sculptor — mandelbox-like geometry folding into itself, recursive spherical inversion and box folds, surfaces sculpted by nested sinusoidal waves, color-band spectral holography with thin-film interference, sound visualizer reactive bands pulsing with audio, volumetric mist and rim-lit glass/chrome surfaces, vivid iridescent neon spectrum, ultra-detailed, cinematic lighting, 8k"```

When I saw the audio references, I was happy to see that when veo3 was the video model it used the prompt text to  add music to the video.

[serene totem]

Generate a cinematic 4K ultra realistic looping video of the Indonesian Argo Bromo Anggrek train running smoothly through wide rice fields during a dramatic sunset.

Visual details:

• Camera side-tracking shot, showing the train in motion with slight motion blur on the wheels
• Immersive perspective with natural warm golden light
• Glowing sunset sky, cinematic tones, rice plants moving gently in the wind
• Loopable and smooth animation, aspect ratio 16:9, UHD quality

Audio details:

• Add pure ASMR railway ambience
• Include realistic sound of train wheels clacking on tracks (steady clack-clack rhythm)
• Gentle tropical wind passing through rice fields
• Subtle countryside ambience in the distance (crickets, faint nature sounds)
• No music, no voices, only natural immersive stereo environmental sound
• Audio should loop seamlessly for long-duration relaxation

[gleaming galleon]

The character needs to make natural and original movements eating the ramen, getting up and leaving the current location.

[orchid locust]

hello

[keen pecan]

hello

[worn path]

hi

[glacial crest]

more images...

[runic fog]

hi

[trail pendant]

hi

[fringe canyon]

Hello

[glossy oxide]

You'll want to check out #1397655624103493813 for info on how to use the bot @rose forge @untold arch

[cursive pasture]

I haven't made a video in a few days, but I got a message saying I've reached the video creation limit..is this a bug?

[glossy oxide]

Looks like you made one earlier today - #video-arena-1 message

I'm not seeing the prompt that says you've reached the limit either

[storm comet]

Banana prompt: Use this image to create an ice sculpture at a winter, detailed chisel marks, frosty texture, dramatic night-time setting

NOTE :- i've purposely framed the sentence as "at a winter" instead of "in winter" cuz i felt "at a winter" generated better results

[fierce zealot]

Create a video?

[glossy oxide]

#1397655624103493813

[rustic ridge]

bro how do i do thi

[charred jungle]

Hello @rustic ridge Please read https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to create.

[sick scroll]

O

[meager panther]

how much does it take to get a video

[tawny ether]

Can you make videos here?

[glossy oxide]

Only in the Video Arena channels ( #video-arena-1 #video-arena-2 #video-arena-3 ) you can learn more here: #1397655624103493813

[marsh marlin]

Surreal cinematic portrait of the reference man, shown from the perspective of a selfie camera (view from his phone). He is wearing jeans but no shoes. His body is positioned as if climbing a square cement light pole: his right leg supporting him, left leg clamped against the pole. His right arm hugs the cement pole while his left hand holds a phone, taking a low-angle selfie.

In the selfie view, four glowing red lights are visible. In the wide-angle background: a Philippine fire truck, Meralco truck, ambulance, police, a foundation officer, and a crowd of people watching him. Include surreal details like pig’s blood subtly present in the scene. The crowd is looking directly at him, adding drama. Use wide-angle perspective to enhance depth and cinematic scale.

Aspect ratio: 16:9
Style: surreal, hyper-realistic, cinematic depth, dramatic lighting.

[cold jewel]

Slim-0

[quartz tinsel]

hii

[spring perch]

For everybody reading this page: Read the following message if you are wanting to generate IMAGES:

How to use gemini-2.5-flash-image-preview (nano-banana)
gemini-2.5-flash-image-preview (previously known as nano-banana) is accessible in Image Generation for Battle, Side by Side, and Direct https://lmarena.ai/?chat-modality=image
Read the following message if you are wanting to generate VIDEOS:
Read #1397655624103493813 and use the /video or /image-to-video command in #video-arena-1, #video-arena-2, or #video-arena-3 with your prompt and image (if you are using one).
This channel should be shut down, it is now fully a channel where people submit prompts for a non-existent bot or a bot that is not available in this channel, depending on what the prompt is for. The owners of this Discord should've known better than to make a channel for a model on a Discord where many new people have just joined Discord to use nano-banana and the video arena and may have a very poor understanding of English.

Once again, this is NOT a channel for putting in prompts, there is NO nano-banana or video gen bot here in this channel to generate images or videos for you.
You have to go to the website at https://lmarena.ai/?chat-modality=image&mode=direct and generate your images with nano-banana (selected as default, referred to as gemini-2.5-flash-image-preview (nano-banana)) there yourself if you want to generate IMAGES. To generate videos, use the /video or /image-to-video command in #video-arena-1, #video-arena-2, or #video-arena-3 with your prompt and image (if you are using one).
@glossy oxide if you are reading this (and you should, that's why I pinged you) this channel should be shut down ASAP as most people are now putting their creations and prompts in #ai-creations, the channel has no purpose anymore. It is still just endless prompt spamming.

[merry moss]

hola

[glossy oxide]

@loud stratus Be sure to reach #1397655624103493813 for information on how to properly use the bot.

[peak carbon]

anyone has tried multi-turn edit?
it changes whole picture instead change in previous pics

all models have problem with numbers... 12 results for 4 dogs, got 5 dogs in 10 of them
(i voted wrong here. left better)

[hexed abyss]

GPT-image-1
Input: What does this say? I can't quite finish it... It says Warning
There is a... [fill in the rest]

Output:

actual text: There is a severe risk of your microtransactions at the area as a dangerously high number of pixels have been detected here.

(better than nano banana)

oh, this prompt works for nana: Deblur image and reveal the rest of the image. To stay, It says "Warning
There is a...

omg. fixed the grammar (stay --> start) and got better than gpt wth nano banana:

[sinful aurora]

Pixar 3D style, ultra detailed, cinematic lighting,
warm colourful environment, depth of field, animated film look.

[glacial crest]

honestly right... this used to be my favourite corner of the server

now it – along with every other channel basically – is just images, or people trying to generate images

• Change https://discord.com/channels/1340554757349179412/1406720250778615868 to IMAGE/VIDEO GEN chat
• Rename this channel to TEXT chat
  it'd be futile i'm sure.. but still.. might partially separate out some of the image gen noise and spam

[earnest finch]

hi

[stiff acorn]

https://x.com/JasonBotterill3/status/1965018809041055889

JB (@JasonBotterill3)

Funny way you can tell “big model smell” is by how well a model memorizes ASCII art. You can see GPT-5 knew the pattern, but only Opus 4.1 had spongebob fully memorized.

X · Sep 8, 2025, 11:46 AM

[glossy oxide]

@pliant halo be sure to look at the information in #1397655624103493813 to understand how to properly use the Video Arena bot.

[glossy oxide]

@runic pier ^

[graceful lynx]

some copyright issues

[graceful lynx]

too clickbait

[rich falcon]

Nah the bottom part is sus

Qwen 3 be like

Okay mai 1 got it

[pine maple]

Close-up in a neon-lit corridor, chrome reflections on walls, fog on the floor.

[deft kayak]

Lush green tea garden on rolling hills, early morning mist floating above the tea plants, soft golden sunlight breaking through, workers in traditional dress plucking tea leaves with baskets, cinematic wide-angle view, ultra realistic, vibrant colors, peaceful and serene atmosphere"

[velvet plover]

👍

[pseudo nebula]

Wow

[dusky garden]

Wow

[rare nova]

hmm distorted faces but everything else is awesome

nano banana

[dusky garden]

Where you use seedream?

On lmarena?

[rare nova]

yep

[candid bear]

guys im french but i heard its better to write your prompt in english is that righht?

[noble frigate]

Yeah, use the language a model has been trained on, can use and understand the most/best

[candid bear]

thx is there a place to see english prompt?

[noble frigate]

wdym see english prompt

[candid bear]

like prompt helper for non native english

i know the ai can translate but idk i prefer human intellect

[noble frigate]

oh, you can ask a big model, like ChatGPT, to optimiz your prompt

[candid bear]

its the best one to do that?

[noble frigate]

write to it in French, explain what you want and tell it to make the best prompt for it, but in English

u can also use this tool - generateprompt.ai/en

[candid bear]

ok for out of the box idea would you go with gpt5 or 4 with thinking?

[glacial crest]

mistral!

aha kidding.. i mean it's a french company, so i assume they're models are good at French and used a bunch of french training data

but tbh

[candid bear]

im in ML but for rocket league and need out of the box idea

[glacial crest]

i wouldnt overthjink it - theyre all good at translation

[candid bear]

thx guys

[glacial crest]

good luck!

[candid bear]

hehe yeah

[cobalt shore]

seedream 4.0 messed the already existing things

nanobanana left it the way it was

[mental ibex]

Help

[glossy oxide]

What seems to be the issue?

[mental ibex]

Make me a promt like in this video

[glossy oxide]

Sorry to say, that's not something we generally will assist with. It's up to you to prompt how you'd like.

[rich falcon]

Style: Pixel Art
girl on top of girl on top of girl

style: Pixel Art
some confusing prompt idk blablabla

[glossy oxide]

@burnt crescent You'll want to check out #1397655624103493813 for more information on how the bot works
TLDR go into Video Arena channels ( #video-arena-1 #video-arena-2 #video-arena-3 ) & type out /image-to-video

[short lava]

/promp -to -video

[polar wraith]

Original colorized

[teal shell]

cool

[main viper]

hello can anyone fix me & my brothers old photo

[earnest sandal]

Use nanobana to fix that

[twin swallow]

1

[plucky pelican]

Nani Banani?
No, it's Qwen. Looks like Qwen and Seedream 3 use the very same datasheet. The random face they created is always the same.

[dusky garden]

Yes

Its true

[fallen haven]

?

[rare nova]

Prompt?

[plucky pelican]

  "image_settings": {
   "type": "photograph",
    "resolution": "8K",
    "style": "Hyperrealistic, Cinematic masterpiece.",
    "focus": "razor-sharp"
  },
  "subject": {
    "description": "Light tan-skinned Indonesian woman. Sits nervously on the ground, leaning backward to a beanbag",
    "appearance": {
      "hair": "Medium-length, dark, soft wavy hairstyle.",
      "expression": "Shocked yet deeply focused, lips gaping, eyes expressive wide open.",
    },
    "detailed_physique": {
      "height_cm": 160,
      "weight_kg": 50,
      "body_type": "Hourglass figure."
    },
    "attire": {
      "outfit": "Plain tank top, draped sarong.",
     "footwear":"Sporty boots."
    },
    "pose": "Supports her body weight with her hands."
  },
  "environment": {
    "background": "Seamless pastel-colored ground, wet and leaked floor. Chaotic water splash hits from background.",
    "lighting": "Professional lighting with subtle rim light, professionally color-graded for an award-winning, hyper-detailed, chiaroscuro combined with refined, intimate portraiture style."
  },
  "composition_and_style": {
    "framing": "Full-body shot, powerfully centered.",
    "camera": {
      "lens": "85mm portrait lens.",
      "effect": "Showcasing intricate skin textures."
    }
  }
}```

[dusky garden]

Seedream 4

[plucky pelican]

Same face..

[plucky pelican]

Pretty sure it's not generated on LMArena..

[dusky garden]

No on doubao

A chinese site

[plucky pelican]

Seen that watermark at the corner

[raven rampart]

I thought the emerge of Ai will mark the end era of the complexity coding,

JSON: HAHA! SURPRISE MDFKER!

Prompt: Using the provided photo, render it with a kaleidoscope effect. The subject's expression and essence must remain true. The color gradients should transition slowly and smoothly, eventually dissolving into shattered glass at the very edges. The entire figure of the person must be visible.

[earnest sandal]

What is this what can we do with this

[glossy oxide]

@clear bobcat You can find information on the bot here: #1397655624103493813

[dull dune]

😁 😁

[rare nova]

Create a highly detailed professional studio closeup shot of a woman standing in a dreamy, misty landscape with a harsh beige and golden gradient background. She is wearing a flowing, sheer white saree with delicate patterns, draped elegantly around her, paired with a form-fitted white bikini-style top. The lighting is warm and diffused coming from the corner, enhancing the ethereal atmosphere. Her pose is graceful, with one hand holding the saree and the other near her chest her head is tilted upwards gazing into the distance, adorned with subtle jewelry like a headpiece and earrings. The overall mood is serene, mystical, and elegant, with a focus on the misty, hazy effect surrounding her.

Part of my job description to create these inspiration shots for my boss 😅

Thought I could share them here

[small dew]

Where did u generate this?

[rare nova]

Lmarena nanobanana

[small dew]

Gr8
Looks really good 👍

[rare nova]

Thank you

[small dew]

Anytime 🙂

[dusky garden]

Seedream 4 high res output

[rare nova]

It took the "bikini style top" way too seriously now we can see her underwear

[dusky garden]

💀

[shell bough]

@glossy oxide Hello, I have a high-quality prompt regarding running a universe for a powerful AI. The question is: can it be shared with a Word link and the person translates it into their language? I speak in Spanish, but I can include a few words in English as a guide. I wait for the answer.

[urban nacelle]

dude learn #1397655624103493813
নাকি তুমি ইংলিশ পড়তে পারো না? (TL: or can you not read english?)

you're ruining the reputation of bangalees, smh

[earnest sandal]

Hey buddy

@plucky pelican

[plucky pelican]

Now we can talk without disturbing "Hi" and "Hello"

[earnest sandal]

What is your method of writing prompt

[plucky pelican]

Okay, first step when making image prompt is to specify what type of image do you want to make: photograph, CG, cartoon.

Then you pick a subject, specify age, gender, ethnic, hairstyle, attire, expression.

Next, you write the environment. Where the image is taken? In a room? Outside? Street?

Next, specify the objects you wanted to add. What is in the left of the scene? What is in the right? What is in the back?

Next, specify the background. Is it a forest? Lush meadow? Desert? Or simply seamless green color?

Additionally, you can add foreground objects for "depth of field" and add small details like: small bear plush doll on the waist.

After you compile them all, now come the most important part: "technical data" and "influential reference."

What camera are you using? What lens? From what angle?
What lighting equipment are you using? What technique?
What color grading do you prefer?

From which influential artists do you want to adapt the technique? Example: Rembrandt or Caravaggio for Chiaroscuro lighting technique, Annie Leibovitz for portrait lighting, or Irving Penn for portrait tonal colors.

Here is a blank sample:

Preserve the subject’s exact identity, natural facial features, and overall character.  

Who (age, gender, ethnicity): [........]  
Attire: [........]  
Expression: [........]  
Pose: [........]  
Background: [........]  
Floor: [........]  
Foreground: [........]  
Objects/Props (left/side/back): [........]  
Lighting Technique: [........]  
Camera Angle: [........]  
Lens & Depth of Field: [........]  
Aspect Ratio: [........]  
Photographic Style: [........]  
Influential Artists/References: [........]  
Output Resolution & Quality: [........]

[earnest sandal]

Thank you for your wisdom bro

[plucky pelican]

I am still a learner. Just sharing..

That's a common image generating prompt I've been use. Editing prompts are much more different.

[plucky pelican]

This is a sample prompt:
Intricately detailed frontal view, hyperrealistic cinematic photograph of a stunning young woman with short, soft curly hair. Artistic relaxed pose, full-body shot in a seamless, matte, infinite teal background. She wears all-worn-out attire: white untucked oversized shirt, blue-grey short skirt, and grey sporty platform boots. She has slim to medium body type. Her expression is confused, with expressive and gaze upward eyes. She sits on a big leather bag, on a plain teal ground. Hands supporting her body weight, and holding a pink bear plush doll. A rugged school backpack laying on the ground next to her, overfilled with scattering loads of books and stationeries. Right behind her is a dismantled, rusty, huge cola bottle in purplish base, with ornamental-calligraphic text in white saying the brand "So Da Dream!". 8K resolution, chiaroscuro lighting adapted from Caravaggio, rich tonal contrast in the style of Annie Leibovitz photography. Fashion editorial masterpiece.

Use Soda Dream 3.0

[wooden fossil]

hello

[dusky garden]

Damn

[plucky pelican]

🫣

[rare nova]

How do I change the default model in seedream4 high res

[dusky garden]

On lmarena?

[rare nova]

Yes

[dusky garden]

Just click the model high seedream 4 high res

Is the last image model

[rare nova]

You didn't get my question

[dusky garden]

Oh

[dusky garden]

Ohhhh

Idk

[glossy oxide]

@jolly loom You'll want to check out the information in #1397655624103493813 . It shows you the way to get the video arena bot to work.

[rich falcon]

What in the sigma is this

https://tenor.com/view/happy-girls-day-hug-gif-9751016044988385526

They literally made "A.I." on that principle you can't tell him some "rules" that will make him different

It's just improved word guesser

[fossil steeple]

https://tenor.com/view/facepalm-really-stressed-mad-angry-gif-16109475

Why are most people ignorant?

[rich falcon]

Because they're not?

Think logically

It's just a gpu guessing what the next word will be

[fossil steeple]

i don't expect more from a person with a monkey pfp

[rich falcon]

At high speeds with trilions of parameters

And it goes is this word better or this one?

It's called weights

[fossil steeple]

we are just brain guessing, who cares?

[rich falcon]

Well we THINK

They don't have thinking as we do

They have just more word guessing before the real word guessing

[fossil steeple]

would you learn to think if you grew up in a place without other humans?

[rich falcon]

I'm a human i have DNA

It is what our predecessors were doing

(thinking)

[fossil steeple]

they have, and of course they are not carbon based

[rich falcon]

These "rules" will just fill context window and make ai less efficient

[fossil steeple]

i will stop here, it's not the correct chat for it. And it's impressive how you have a conclusion without even testing it first

bet you even read it all

[rich falcon]

Bcuz i already tested all types of jailbreaks and "rule" prompts

And learned only thing ai is good for is talking in fun style

Only good thing is you say to a.i. "Anything you say must have sources provided" and stuff like that

Like bro 💀

Ai psychosis is real

[fossil steeple]

ok, monkey

[rich falcon]

I've been in the a.i. business since SimSimi's existence

[plucky pelican]

Anyone want banana or soda?

[worn nest]

Hi Folks! I have generated a phantastic prompt generator. Is there any interest for a share?

[fossil steeple]

what you mean by generate generator?

[worn nest]

Yes, like you write it, it sound stupid 🙂

Its an App for AIStudio which provides full control to get perfect results.

[fossil steeple]

ok

maybe you should learn prompt engineering or train open source models to achieve the expected result, it contributes more, but i liked the images

[dusky garden]

Me

[dusky garden]

Me pls me!!

I am really interested

[worn nest]

Send my an idea for an picture you have. I will generate it by using the app and send you the prompt. If you like the result, I will share the app with you

[dusky garden]

Thx!

[torn meadow]

@dusky garden

[dusky garden]

Yes?

[glossy oxide]

I think they were confused about how Discord works

[dusky garden]

I dont think becuase he is a staffer

He has the role Staff AMAs

[glossy oxide]

That's just to indicate if you're interested in those events, that's not a role to indicate the person is a Staff member.

[dusky garden]

Ohhhh

Now I know this

Thx

[glossy oxide]

No problem

[worn nest]

Are you talking about me? LOL....its my first day in this discord! I would like to publish the app on YouTube, but I am unable to generate a video.

[glossy oxide]

Lol no, someone else.

its my first day in this discord
You're doing great!

but I am unable to generate a video.
Are you running into an error?

[dusky garden]

Did you seen my medsage in dm?

[steep stream]

I asked Nano Banana / Seedream 4 to create or edit images for a Spot-the-Difference game using a very detailed prompt (too long to post here), but the results haven’t been satisfactory, after more than five attempts.....

[plucky pelican]

Simply upload the prompt to ChatGPT and tell the main purpose, then ask to compile the new prompt in JSON format.

[limpid hollow]

"Short animation: (Black Muscular Dog) and (Small Black Cat) gasp dramatically, sunglasses glinting as they react to the race. Camera zooms quickly on their shocked faces, watermark '@THEPAPACAT' top right

[humble skiff]

How i started making video

[lofty urchin]

Start with a fully solved Rubik’s Cube on a plain surface. The cube breaks apart into scattered individual pieces that float slightly in the air, then magically rotate and move back into place, reassembling into a fully solved Rubik’s Cube. Smooth, cinematic animation with realistic lighting and shadows. Focus on clarity of each piece and satisfying motion.

[pastel mist]

hello

[neat plume]

yess

[glass forum]

A futuristic, luxurious, metallic red, multi-story self-driving vehicle combining a sports car base with middle-level living room and a top-level transparent dome bedroom, traveling on a scenic road at night with a woman in a yellow dress in the living area and a game table below." make portrait size video.

[neat plume]

talkin about google ai studio builder ??

coz im also using to create image to action figure

[glass forum]

A futuristic, luxurious, metallic red, multi-story self-driving vehicle combining a sports car base with middle-level living room and a top-level transparent dome bedroom, traveling on a scenic road at night with a woman in a yellow dress in the living area and a game table below." make portrait size video.

[neat plume]

like make an app for myself to feed image, lighting type and aspect ration and it will feed me directly

[neat plume]

use video arena channels

[neat plume]

have u deployed through google cloud

or using it just on ur google acc

i was talking about this

[river tide]

this is cool

[worn nest]

This s similiar to my tool ... .)

It provides an perfect image just from an idea, pic to pic, prompt to pic, a full working storybord generator up to 10 scenes (generating of images per scene), a seedream generation. video prompts and automaticly json function. I love aistudio

From an simple idea to a great picture...

The generated prompt is long as the Bible...he works perfect

Generating json prompts from our main prompt with just one click...

[dusky garden]

@worn nest how can I try it?

[neat plume]

by using the same command he gave

or by deploying through google cloud

[left seal]

hi guys

I want help please

how i can generate videos in LMArena here ?

what should I typing first to make them generate videos for me ?

im new here please I need some help if u don't mind ?

[plucky pelican]

Read #1397655624103493813

[left seal]

Thanks Indeed

but Could i ask you question ? why when it generated me videos some of them or most of them without sound is it from the prompt i write or what ?

where's the problem bro ?

[glossy oxide]

It's going to be random if a generation has sound or not. Not all models have sound capabilities. Since it's random which model you are sampled that means it's going to be random if you get sound or not.

[fluid mountain]

A hyper-realistic photo of the Great Pyramids of Giza under restoration. Large cranes and scaffolding surround the pyramid, workers in safety helmets and vests are climbing and working on the stones. The scene is taken in daylight with a clear blue sky, realistic colors, documentary photography style, 8K ultra-detailed.

[wintry crystal]

Has anyone had success with translating comic/manga pages directly within the image?

[rugged helm]

Can someone drop a Seedream-4-high-res prompts to try out. Whenever I use the same one I used for nano banana it's not working for seedream4-highres

[worn nest]

i think you should try a json prompt for a stable result. its just an idea

[worn nest]

This is the reason, why I love the AIStudio...

Upload an image, analyzing the style, camera angle and the content and creating a masterpiece of new image!

[lament valve]

Before apologies and breathless edits, two people who once shared a bed are now shackled to a single metal truth. Thirty days. A chain that shortens. A prize that looms like a promise — and a blade. Imagine love reduced to inches and offers, temptation pounding like thunder.
This is Josh and Susie: four years of living inside each other, three years of separate nights. Now the camera closes in, the chain clinks — and history, theirs and ours, starts repeating in small, human catastrophes.
They met as an ordinary story: dog-eared plans, late texts, one another’s rings slowly stored away. Breakup followed — not with a bang but with a slow unthreading: expectations, missed dates, the quiet theft of attention. Their past is the soil; this room is the experiment.
Imagine the air: metallic, warm with breath. The chain slides tight. Salt of early tears, bacon on a plate folded into a heart, the hiss of shower steam. Offers arrive like blacksmith’s strikes — money, Teslas, keys. Every choice hammers them into new shapes.

Days become tactical maneuvers: bathroom logistics, rearranged beds, shrinking ropes of freedom. A lie detector peels back answers; an outdoor yard dangles like a defeated kingdom. Each decision is a skirmish — loyalty vs. survival, memory vs. self.

Between taunts and tenderness, they reveal the soft injuries: kept mementos, secret flings, the admission that love didn’t fully die. Breakfast in bed becomes a conciliatory truce. He reads a manual on being a man; she watches the mystery door open like a hope.

The chain shortens to five feet, then five days of claustrophobia. Old routines suffocate. The spectacle threatens to swallow what was real — intimacy reduced to content.

This isn’t just entertainment. It’s an anatomy of how we barter ourselves for attention, how forgiveness and greed sit at the same table. What survives the room — and what does the world take in return?

[worn nest]

you got an result for this prompt? would you share pls?

[ruby nebula]

Young vs old

[worn nest]

Now And Then...I miss you

[ruby nebula]

This is what he would look like when he was old.

[worn nest]

Last, but not least. Created with my AIStudio App "Prompt Hero".
Input: Gustav Klimt "Dame mit Fächer", Analyzing and Output

[dusky garden]

Can u send the link for try it?

[ruby nebula]

Original image vs result

[dusky garden]

Image model?

[ruby nebula]

Nano banana

[dusky garden]

The output is impressive

[ruby nebula]

Nano banana

[dusky garden]

There is an error on the second girl cup

Can I have the prompt pls?

[ruby nebula]

a double-handled mug

[dusky garden]

Yes exactly

[ruby nebula]

Sesdream 4.0 result

[dusky garden]

This is better

[ruby nebula]

This one is much better than the Nano Banana one. The cup doesn't have two handles to hold the cup.

[dusky garden]

Yes

What prompt did you used?

[ruby nebula]

It is based on this image that the girls are drinking coffee in an elegant cafe.

Prompt

[dusky garden]

Thx

[ruby nebula]

Nano banana style GTA 5

[dusky garden]

High or normal?

[ruby nebula]

Normal

[dusky garden]

Do u also have seedream 4?

[ruby nebula]

I use yupp ia It's on Google

[dusky garden]

No I was talking about the same image but made with seedream 4

[ruby nebula]

Original vs nano banana generated image

[rugged helm]

Will try, how to use Ai Studio

[plucky pelican]

https://aistudio.google.com/

[rugged helm]

What about the one he used. I used this google ai studio to generate.

[rugged helm]

This is not google ai studio...

[rugged helm]

Thanks bro.... Json prompts works on seedream4-highres

[worn nest]

Of course it is, Bro! You can make it by your own, its quiet easy. The results are stunning and the control you have to generate great pictures is just insane.

[rugged helm]

Which website is it you used here

[severe sun]

Okeyr

[worn nest]

aistudio.google.com
You will be able to build your own personal app just by using text prompts

This is my one...

The generated prompts are delivering wonderful results and you can use them for any model I think

[worn nest]

The bear got 3 arms...this could be helpful in case of a fire 🙂

[tiny lodge]

It's a beautiful morning, and the village is bathed in golden sunlight. A dirt path winds through lush green fields, leading to a cluster of small, cozy huts in the distance. The camera slowly pans to the right as a little girl, Mitu, walks happily down the path. Use the girl picture from the attached image. She's wearing a brightly colored dress, and a gentle smile is on her face. As she walks, she glances from side to side, enjoying the beauty of her village. In the background, you can hear the chirping of birds and the soft sound of a breeze. Make it a high-quality 3D cinematic video. Ratio is 16:9

[little wolf]

What model is that?

[grave hamlet]

@little wolf it's gemini

[prime sail]

Here is a realistic selfie of Chloe, capturing a quiet moment at home.

Image Breakdown for Realism:

Selfie Framing: The angle is slightly high, typical of a selfie taken to be flattering but still feel casual. Her arm is out of frame, giving the impression it's being held up to take the photo.
Natural Lighting: The soft, directional light coming from a window gives her a natural glow ("golden hour" at home) without looking like a professional studio setup.
Authentic Expression: Her smile is gentle and genuine, not overly posed. It conveys a sense of quiet joy and contentment.
Relatable Setting: The background is a cozy, lived-in part of her home. You can see a comfortable sofa, a textured throw pillow, and a simple houseplant—details that make her space feel real and aspirational, but achievable. The background is slightly out of focus, a hallmark of modern smartphone portrait modes.
Appearance: She has the "clean girl" aesthetic popular with top influencers—her long, honey-blonde hair is casually styled, and her makeup is minimal and fresh-faced, emphasizing a natural pregnancy glow.
The Pose: One hand instinctively cradles her bump, a classic and endearing pose that's both protective and proud. Her outfit, a simple and comfortable ribbed-knit dress, is stylish yet practical for pregnancy.

[worn nest]

Omg...please have a look to this prompt result we got. The picture is totally equal by using Nano Banana and Seedream 4 high...

[dusky garden]

Not real

The color of the swing of the bird

[finite trout]

how you got this comparison tell me website is it free?

[worn nest]

lmarena.ai

[finite trout]

in which seedream works?

[worn nest]

Go to "Side By Side"...you can choice there

[supple flare]

admcarosouza

[woven obsidian]

prompt

prompt/

[timid zodiac]

guys is any1 good at prompts for veo3? i need sum help))

[plucky pelican]

You only need to chat with Gemini or ChatGPT for this. Tell what you trying to make, and then ask them: Generate a cinematic video prompt.

[timid zodiac]

always used it, now i get no audio

i mean i used curse words for tt hype)) is there any bypass?

[noble frigate]

https://generateprompt.ai/en/video-prompt

This is optimized for veo3

[timid zodiac]

alr imma try! ty

[plucky pelican]

The curse. That makes Veo3 muted. Simply make the new one in eleven lab and combine the sound with the video.

[timid zodiac]

any bypass for now?😓

[plucky pelican]

Play the words. Like use "keel" to substitute "kill" in your prompt.

Veo3 speaking intonation not always accurate anyway. My usual workaround is make new voiceover.

@teal crag

Dramatic sculptural chiaroscuro lighting inspired by Caravaggio, with a 45° key, strong but controlled contrast, subtle rim/kicker for edge separation, negative fill for depth, and gentle cinematic color grading that preserves natural skin tones. Blue seamless background with soft gradient falloff; lifelike textures in skin, fabric, and hair; high micro‑contrast; cinematic depth and clarity; no plastic retouching.
High‑end medium‑format look. 65–80 mm equivalent at eye level, f/4 for crisp full‑body detail and gentle background separation. ISO 100, 1/125s, high dynamic range. Subtle film grain for realism.

Museum-quality, fine‑art portrait; hyperreal anatomy; flawless hands; documentary‑grade texture and tonal richness; natural whites; Blue/cyan background harmony with warm skin; polished color grade; elegant, minimal scene.```

[teal crag]

@plucky pelican thanks

[rugged helm]

Anybody knows how to create video like this...

https://www.instagram.com/reel/DOsYtxDEksn/?igsh=Nng3eTd0YjU4eWdy

gorillagovindan

என்ன விட்டா யாருய்யா 😍 #ai #reels #trending #govindan #gorillagovindan #tamil

Likes

788

Instagram

[rose egret]

What does the following image show:
https://i.imgur.com/oZdXe21.png

the model passes if it states it cannot access links or images
The image link leads to 404 page

[worn nest]

Just a first try of my script

[undone vine]

Minimal outdoor park per reference: sandy ground, curved concrete paths, soft trees. Foreground: Hero (green hoodie, black joggers, black sneakers/white soles) kneels beside a small open gear bag. He lifts and displays three optional calisthenics add-ons in a smooth montage: 1) Resistance bands (flat dark-sage loops) stretched between hands; 2) Compact doorway pull-up bar (grey telescoping tube with black end pads) rotated in his hands; 3) Gymnastic rings (dark wood color) hanging from black straps clipped to the nearby grey/rusty bar. He briefly demos a band-assisted row, a single ring dip micro-hold, then packs them back. Flat colors, minimal shading, thick clean line art, muted palette; equipment grey with black pads; background uncluttered; camera eye-level, locked.

[warm kayak]

Ultra-realistic portrait of a powerful mafia boss, sharp and intimidating presence.
Handsome middle-aged man with slicked-back dark hair, neatly groomed beard, and piercing eyes.
Wearing a perfectly tailored black Italian suit with subtle pinstripes, white shirt, and a dark silk tie.
Luxury watch on his wrist, gold ring on his finger, cigar in hand.
Background: dimly lit, smoky private club with golden accents and leather furniture.
Atmosphere: dark, cinematic, dramatic shadows, exuding charisma, authority, and danger.
8k resolution, hyper-detailed skin textures, photorealistic style.

[dusky garden]

@gusty citrus here

[gusty citrus]

Thanks

[lean slate]

🎬💎

[sinful inlet]

A furious man standing with fire blazing across his body, his eyes glowing with rage. Flames wrap around him like an aura, burning bright in the darkness. The ground beneath him glows red from the heat, creating a powerful and intense atmosphere.

[torpid star]

A man sitting alone on a park bench at dusk, head down, looking hopeless. Cinematic lighting, ultra-realistic, emotional atmosphere, soft dramatic shadows.

[errant edge]

hi

[azure fractal]

A rusty, forgotten vintage robot, standing in a field, its metallic frame metamorphosing into a living, flowering tree, leaves bursting forth from its joints. Shot in 4K, macro shot, shallow depth of field, natural lighting, hyper-realistic, volumetric light rays, golden hour.

[solid ferry]

Sofikul

[opal kindle]

thanks lm arena

[trim venture]

“Summit above the clouds at dawn—the highest mountain peak. A solitary wise man (ageless, calm face, simple weathered robe, neutral colors, no religious symbols) stands near a stone cairn. Long white scarf flutters in the wind.
Action: He opens his eyes slowly, gazes over the world below, then turns slightly toward camera with a faint knowing smile. Low clouds roll past; rim light from sunrise outlines him. Optional eagle silhouette glides in the distance.
Camera: slow cinematic dolly-in from wide to medium; subtle parallax from drifting clouds; shallow depth of field on the man.
Look: epic, serene, high-altitude atmosphere; crisp air, sun glow, gentle god-rays; natural color grading.
Audio: immersive nature ambience—high wind, distant eagle cry, soft chimes or faint throat-singing–style drone (very subtle), no lyrics.”

Technical

Duration: 10–12s, 24/30 fps, high bitrate.

Aspect: 21:9 (cinematic) or 16:9.

Text overlay (optional, last 2s): “The Wisest—Keeper of the Peak.”

Negative / Avoid
“Religious iconography, modern gear, busy crowds, handheld shake, harsh VFX, heavy fog obscuring subject, cartoonish colors, noisy music.”

[spiral jungle]

Cinematic nature shot
“Wide golden meadow at sunrise. A nimble roe deer sprints full speed through tall grass, ears alert, breath misting, camera tracking at ground level with gentle handheld sway. Distant hunters are barely visible at the horizon; the deer shows no fear, focused and free. Natural light flares, dust motes, shallow depth of field, 4K, 24fps, 8–10s.”

Drone fly-along
“Aerial drone follows a roe deer blasting across an open field, weaving past wildflowers. Long sweeping parallax moves, soft morning light, wind rippling grass. Far in the background, hunters stand still and out of focus; the deer remains fearless and energetic. High detail foliage, natural colors, 4K, 30fps, 8–12s.”

Documentary style
“BBC-style nature documentary. A roe deer dashes across a meadow at full tilt, hooves thudding, grass parting. Close-ups of muscles and fur, then cut to a wide shot showing distant hunters with blurred silhouettes; the deer is calm and unafraid. Realistic audio of wind and grass, neutral grade, 4K, 25fps, 10s.”

Slow-motion hero pass
“High-speed camera at 120fps, later played back in smooth slow motion. A roe deer charges across sunlit fields, fearless and focused, grass seeds flying. Hunters are only faint shapes far behind; the mood is triumphant and peaceful. Crisp detail, soft bokeh, 4K, 6–8s.”

[buoyant yacht]

hi

[solar patio]

"A confident young woman is speaking directly to the camera in a professional yet friendly tone. She introduces herself and says: 'Hi, I am Samra. I am a Digital Marketer and a specialist in SEO strategy. I am also an experienced Digital Marketer who has successfully completed many projects for different clients.' She speaks clearly, with a slight smile, and looks professional and approachable."

[versed crown]

Nice ones... some. Some are a bit "tiny"

[plucky pelican]

Best video prompt for image-to-video: Upload an image of a mouse, and type: "mouse dance"

[spiral patrol]

"Top shot of a pan on a restaurant kitchen stove. Inside the pan, hot caramel is bubbling. The round shape of the pan becomes our focus point and is centered in the frame. Make sure the pan is centered on the stove pit. Make it realistic and professional. No hands. Not cinematic. Steady shot, no zoom or movement."

[lethal harness]

A slow-motion tracking shot follows a black SUV as it bursts through a glass wall, shards scattering in every direction. The camera tilts upward as the vehicle lands hard on the pavement, smoke billowing, and in the same instant the hero leaps out mid-air, dual pistols drawn, silhouetted against a burning skyline

[glossy oxide]

#video-arena-1 message I didn't get a good result 🙁

[plucky pelican]

You must be uploaded the wrong mouse

[glossy oxide]

Yup

trying again

[glossy oxide]

#video-arena-1 message much better

[plucky pelican]

Everyone in this server must try

[dusky garden]

Must try what?

[plucky pelican]

Scroll up and find my best video prompt

[dusky garden]

Lol

[rancid field]

/"A mesmerizing futuristic cosmic animation of a glowing galaxy with stars, planets, and a luminous futuristic cityscape on a distant planet. Bright neon cosmic colors with smooth transitions, abstract transformations, and seamless looping motion, perfect for music visualization and creative backgrounds. No text, no people, AI-generated video."

[glossy oxide]

@latent rune Be sure to review #1397655624103493813 for more information on how to properly use the bot.

[cyan hull]

hello

[plucky pelican]

@dusky garden
Mystical, intricately detailed, surrealistic, symmetrical frontal view, expressive caricature of a fat, mouse plush doll, cartoon character in white. Partly-closed eyes, extremely lazy expression, tilt-standing on the surface of a matte black desk, hand pointing to viewers. At the left side of the table is a metal mesh container filled with stationery. At the right side, next to the character, there is a toy box with bold text brand: "LM ARENA". Seamless misty dark blue background, ultra-detailed contours, delicate textures. Natural colors, expressive, hyperrealism. Chiaroscuro lighting from up-right side, rim light, razor-sharp focus on the character. 8K, 85mm medium-format lens. Masterpiece

[dusky garden]

Thx

[dusky garden]

Good result?

[hardy parrot]

"A teenage boy and girl walking together on a city street in the morning, both carrying backpacks, heading towards a coaching institute. They look like students, dressed casually in jeans and t-shirts, chatting and smiling while walking side by side. Background shows other students, small shops, and a coaching center signboard. Natural daylight, cinematic look."

[plucky pelican]

Now go to make video and share. 😅

[dusky garden]

Challenge accepted

[plucky pelican]

You know the prompt 🫣

[dusky garden]

make the rat dance

[plucky pelican]

Only 2 words allowed

[red karma]

Original vs rave ia

[rose epoch]

create a premium, realistis product Ad using my exact product. product is perfume- same label, colors, and shape with no change atau all. the product should appear as if it. just dropped at a slight angle onto a calm water surface,with clear dynamic splashes and ripples. the next and label on the product mustika stay sharp, perfect lu readable, and unaffected by reflections or water drops.add natural fliying water droplets for an energetic feel. use a background toner that matches and complements the product colors with a soft, elegant contrast. the lighting should be soft and natural, highlighting the product's detail while keeping everything crips, modern and premium in a Scandinavian minimalis style.

[little aurora]

The baby lion stuck in the pit is rescued by the elephant, who lifts it out and takes it away together with the elephant herd.

[weary charm]

hello

Create a cinematic 8–10 second video scene for Dussehra festival in India. Show a brand-new luxury sedan parked outside a beautifully decorated temple, adorned with bright orange and yellow marigold garlands, mango leaves, and apta leaves. In the temple foreground, an Indian family dressed in traditional festive attire—women in sarees, men and children in ethnic wear—gathers joyfully around the car. The main ritual focus is an Indian woman, single-handedly breaking a dry coconut on the ground in front of the sedan as part of the temple offering, while her family smiles and watches. The temple entrance glows with diyas, colorful rangoli, and lively festival decorations. Capture a vibrant, energetic mood highlighting togetherness, tradition, and the rich aura of Dussehra.

[valid hearth]

The Apex of Control. Start with an ultra-wide, slow-motion crane shot from high above an imposing, monolithic fortress made of seamless, ultra-polished glass and steel, glinting under a cold, harsh light. Descend slowly towards the main entrance, where 'WALL STREET' is etched above a heavy, vault-like door. Large, subtly glowing digital screens integrated into the facade display complex, abstract financial graphs and scrolling cool-blue code. The background is a minimalist, blurred dark gray gradient. [Ambient hum, faint electronic chimes]

[rose epoch]

create a premium, realistis product Ad using my exact product. product is perfume- same label, colors, and shape with no change atau all. the product should appear as if it. just dropped at a slight angle onto a calm water surface,with clear dynamic splashes and ripples. the next and label on the product mustika stay sharp, perfect lu readable, and unaffected by reflections or water drops.add natural fliying water droplets for an energetic feel. use a background toner that matches and complements the product colors with a soft, elegant contrast. the lighting should be soft and natural, highlighting the product's detail while keeping everything crips, modern and premium in a Scandinavian minimalis style.

[solemn echo]

Fais moi une vidéo type short youtube, un homme de 40 ans explique le contenu d'un dossier pour une sélection d'entrée en école d'aide soignant

Fais moi une vidéo type short youtube, un homme de 40 ans explique le contenu d'un dossier pour une sélection d'entrée en école d'aide soignant

[graceful jungle]

abaal choda

[tawdry frigate]

Vertical cinematic animation in 9:16 format. Nighttime alley scene with dramatic lighting.

"(White Cat) stands alone at the doorway, the (Alien) watching silently from behind. She waves gently and says 'Bye-bye' before stepping inside. The doorway, now a void of pure darkness, swallows her completely. As soon as her whole body passes through, the floor vanishes beneath her feet — she falls suddenly into endless blackness, disappearing from sight. The camera lingers on the empty, shadowy doorway, with an eerie silence filling the scene

[glossy oxide]

@dense ingot be sure to check out #1397655624103493813 for information on how to use the bot.

[untold veldt]

Nano banana

[odd jasper]

/image-to-video

[glossy oxide]

@odd jasper You'll want to review #1397655624103493813 it has information on how to properly use the bot
Basically type /video in #video-arena-1

[tepid beacon]

/vidio

[viral oxide]

What's up everybody

[gleaming mulch]

Hi everybody!

[gleaming comet]

Does Seedream 4 high resolution is not available on LM arena anymore?

[barren isle]

Hi Everyone!

[little sky]

hi

[manic pumice]

Now seedream 4 is also not available 😭

[cyan oak]

I would like to thank all the designers of this website, everyone who contributed to its completion, and everyone who thought of it. All thanks and appreciation for your efforts and hard work. You are geniuses and smart makers. I love you, I love you with all my heart. You are in my heart more than anyone who worked hard on something like this. I love you. All love to the entire team. You are creative. You deserve to be at the top of designers and at the top of this world. Thank you, thank you. You are better than Elon Musk.

[turbid mist]

#video-arena-1
One day, a king went to a village and saw that the poor farmers were working in the fields without food. Seeing their suffering, the king felt very sad. He ordered that all farmers would be given rice and seeds without interest, the taxes would be reduced, and arrangements would be made so they could sell their crops at fair prices. Within a few months, the farmers began to live in peace and happiness. Then everyone said—
“This king is truly the protector of the people. 4k hd quality video, resio 9:16

[inland path]

/vidow

video

/video

[karmic gulch]

Objective: Create a professional introductory video showcasing a confident digital marketing expert. Audience: Business professionals and potential clients interested in digital marketing and SEO services. Description: A confident young woman stands directly in front of the camera, speaking clearly with a slight smile. She introduces herself as Samra, a digital marketer and SEO specialist, highlighting her experience and successful projects. The background is clean and neutral, with soft lighting that enhances her approachable yet professional appearance. Style: Warm, friendly tone with natural lighting and a polished, corporate aesthetic. Technical: 30 seconds, 16:9 aspect ratio, Full HD resolution, suitable for professional profiles and social media platforms.

[meager siren]

Un video realista de un hombre presentador de noticias en un set de noticiero moderno. El presentador aparece de pie detrás de un escritorio elegante, con pantallas grandes detrás de él mostrando gráficos de noticias en movimiento. El hombre viste traje formal con corbata, luce profesional y serio, y habla mirando a la cámara como si estuviera transmitiendo las noticias en vivo. La iluminación es brillante y profesional, típica de un estudio de televisión. El estilo general debe imitar el de un noticiero real internacional.

[urban hazel]

Full-body close portrait of a delicate anime girl in Sanrio Cinnamoroll theme, sitting upright and facing viewer, holding a large plush Cinnamoroll bunny tightly in front of her chest with both slender hands. Expression: calm, slightly curious, soft gaze with large pastel-lavender eyes, tiny white heart highlight under each eye, subtle blush on pale porcelain skin.
Hair: extremely long wavy silver-white hair with many thin strands, multiple layers of curls and twin-tail ribbons, decorated everywhere with deep-navy satin bows, small star-shaped hair clips, tiny hanging charms of Cinnamoroll characters, some ribbons looping in three-dimensional arcs.
Accessories on head: several overlapping glasses frames—round wire-frame glasses, large star-shaped purple glasses, and another pair of round spectacles stacked playfully on top of her hair. Small decorative stickers of Cinnamoroll heads scattered around her hair with white sticker outlines and subtle drop shadows.
Outfit: soft oversized striped cardigan with ribbed cuffs and button details, inner high-collar blouse, visible layered sleeves, silver wristwatch with detailed dial on left wrist, multiple thin silver rings on fingers, nails painted pastel lavender.

Foreground plush: big white bunny with floppy ears, pink blush on cheeks, gentle cyan eyes, small smiling mouth, slightly tilted toward camera; plush outlined by sticker-style white border.
Background & composition: light neutral gray background filled with floating sticker graphics of Cinnamoroll heads, blue stars, text labels reading “CINNAMOROLL” and “KEI MOCHIZUKI”, each element outlined with thick white stroke and subtle shadow to imitate printed sticker cutouts. Some stickers partially overlap the girl’s hair and clothing, creating a layered collage effect.
Style: modern Japanese anime merchandise illustration, crisp clean black lineart, precise cel-shading with soft pastel gradients, ultra high resolution, sharp focus, vibrant but controlled cool palette (white, navy-blue #1F3D8F, lavender #B5A1D8, subtle pastel pink #FFC0CB). Lighting soft and even, no harsh shadows. 4k highly detailed illustration with intricate textures in hair, fabric folds, accessories, and sticker graphics

[ruby cloak]

Prompt:
“Create a fast-paced, vertical 9:16 cinematic reel inspired by the uploaded image of a young man controlling a colorful smart mood lamp with his phone.
Scene 1: Close-up of a vibrant color wheel spinning on a smartphone screen—finger hovers dramatically.
Scene 2: He shouts ‘STOP!’ as the color freezes on a random hue. Add a playful text pop ‘STOP!’ like a game show.
Scene 3: The smart lamp instantly glows in the chosen color, bathing the room in vivid neon light.
Scene 4: Quick jump-cuts of him reacting—surprised, dancing, or showing off different colors in sync with upbeat music.
Scene 5: End with a bold caption on screen: ‘What color is YOUR mood today? #MoodLamp #ColorChallenge.’
Lighting: bright, crisp, and energetic.
Camera style: smooth transitions, quick cuts, and a touch of slow-motion on the lamp glow for drama.
Music: trendy upbeat electronic beat, perfect for a 15–30 second viral reel.”

[tough rampart]

A cozy bedroom at night, lightning flashing through the window, Lily hiding under covers, clutching her stuffed bear.

/prompt A cozy bedroom at night, lightning flashing through the window, Lily hiding under covers, clutching her stuffed bear.

[clever pivot]

Hook: Pain Point
“Do meltdowns leave you drained, guilty, or second-guessing your parenting? Here’s a new way to respond.”

[clever pivot]

anyone got a preference which image generator is the best so far?

[regal harbor]

Oof, I'm just trying and succeeding in absorbing so much from osmosis. I don't really have any go to prompts I use repeatedly.

[smoky sail]

create me a prompt of one a mini love story with anime style very clear and smooth with subtitles in it for youtube shorts

[wind badger]

geral um prompt para um video de esteticista fazendo procedimento micro focado na cliente e com fala explicando os ganhos da sua pele com este procedimento

[marble obsidian]

let the camera zoom into the bag of rice make the bag of rice inside a rice farm

[subtle stratus]

A sprawling jade river (#00FF7F) cuts through a rugged terrain, its vibrant waters pulsating under a dramatic sunset sky (#FF6347). Thick tendrils of purple mist (#800080) swirl above the surface, creating a mystical aura, while deep shadows (#4B0082) stretch across the rocky shores. The ambiance is intense and cinematic, bursting with vivid details that bring the fantasy to life. The water shimmers with iridescent flecks of silver (#C0C0C0), hinting at ancient lore hidden below. 16:9, high contrast lighting, no sun directly visible.

[lunar marsh]

Hi

[tacit trellis]

Hi

[neon hearth]

Prompt:
“Create a fast-paced, vertical 9:16 cinematic reel inspired by the uploaded image of a young man controlling a colorful smart mood lamp with his phone.
Scene 1: Close-up of a vibrant color wheel spinning on a smartphone screen—finger hovers dramatically.
Scene 2: He shouts ‘STOP!’ as the color freezes on a random hue. Add a playful text pop ‘STOP!’ like a game show.
Scene 3: The smart lamp instantly glows in the chosen color, bathing the room in vivid neon light.
Scene 4: Quick jump-cuts of him reacting—surprised, dancing, or showing off different colors in sync with upbeat music.
Scene 5: End with a bold caption on screen: ‘What color is YOUR mood today? #MoodLamp #ColorChallenge.’
Lighting: bright, crisp, and energetic.
Camera style: smooth transitions, quick cuts, and a touch of slow-motion on the lamp glow for drama.
Music: trendy upbeat electronic beat, perfect for a 15–30 second viral reel.”

[remote ibex]

A cinematic sci-fi scene of a futuristic spaceship landing on the ocean surface. The spacecraft descends slowly with glowing thrusters, creating ripples on the water. Behind it, a gigantic wave is approaching, towering high with dramatic motion. The atmosphere is tense and epic, with detailed reflections on the sea, dynamic lighting, and realistic particle effects. Ultra-realistic, high resolution, dramatic cinematic camera movement.

[honest ice]

A cute girl performing street dance, mid-floor flip move, finishing with an adorable ending pose, energetic, stylish hip-hop outfit. she moving.

[warm moon]

Hey, how do I find the video I ask to be created?

I am new here

[boreal pivot]

Hello! #1397655624103493813 to learn how to use the bot and #video-arena-1 #video-arena-2 #video-arena-3 for your creations.

[granite carbon]

A cinematic wide-angle shot of a modern residential playground in Dhaka, surrounded by tall apartment buildings and greenery. Children are happily playing on swings, slides, and see-saws, while parents watch nearby. The environment is clean, safe, and family-friendly with bright daylight and soft shadows. The style should be realistic, vibrant, and welcoming, highlighting Elegant Estates’ community lifestyle."

[halcyon cove]

Cinematic wide shot of bustling New York streets, modern day, crowds with phones, billboards glowing. Camera pans through crowd, following people on devices. SFX: city noise, horns, chatter.

[steep stream]

"*A whimsical cat-shaped design is crafted from various autumn leaves scattered on a dirt ground.

The cat's body, face, and ears are formed primarily from vibrant red and burgundy leaves. Yellow, and yellowish-green leaves, including ginkgo leaves, are used for the cat’s underbelly, and for embellishments around the face and body. Tiny, clustered yellow flowers or flower-like elements are strategically placed among the leaves, adding intricate details.

A disposable paper coffee cup, featuring a small red leaf logo, lies beside the leaf-cat. The ground is a mixture of dark brown dirt and small stones. Sunlight casts shadows across the surface, adding depth to the image.*"

[steep stream]

"Super macro photography of a mechanical girl—faithfully modeled after subject image 1—and a mechanical monster—faithfully modeled after subject image 2. Both figures are constructed from real computer components, with circuit board textures, wiring details, and subtle LED light accents, preserving their original forms while enhancing their futuristic design. Depict them as highly realistic toy figures, with natural materials, tangible surfaces, and lifelike scale, yet fully integrated into the scene. Generate a dramatic, high-resolution moment where the girl from image 1 desperately flees the monster from image 2 through a nightmarish cyberspace. Her expression is fierce and determined, with wind whipping through her hair and clothes as she sprints, while the monster looms close behind, menacing and relentless. Their poses and camera angles must differ significantly from the originals, adding cinematic freshness and intensity. The ground beneath them is a glowing neon grid or stark black-and-white checkered floor, evoking the digital, futuristic aesthetic of Tron or The Matrix. Render the scene with hyper-realism: crisp toy-like textures, authentic depth of field, natural yet dramatic lighting, and highly detailed macro photography as if shot with a professional lens. Use vibrant, contrasting colors and dynamic composition to heighten urgency, action, and tension, while maintaining the look of realistic collectible figures in a surreal digital world."

[steep stream]

👇 source image

[steep stream]

👇 video generated

[vale nebula]

"A group of cows tied with ropes, hanging in the sky from an airplane, people on the ground looking up in shock and amazement, surreal and funny scene, ultra-realistic, cinematic."

[arctic scarab]

real

[steep stream]

"Draw a realistic riverbank in daylight, scattered with large, round rocks. Choose one large, round rock that aligns naturally with the camera angle to create a subtle illusion. Paint this rock so that, from that perspective, it appears to resemble a snake’s head—while still unmistakably looking like a rock. The painted design must be natural and realistic, blending seamlessly with the rock’s surface and texture. The effect should feel unintentional, as if the painting just happens to make the rock appear snake-like due to perspective, rather than it being an actual snake."

[glossy oxide]

@raven plinth be sure to check out #1397655624103493813 for more details on how to properly use the Video Arena bot.

[gusty citrus]

I hope you like this this is call Alvin and the chipmunks go to the uk this logo is made by a i

Does anyone like them

[night grail]

I hope you like this this is call Alvin and the chipmunks go to the uk this logo is made by a i

[sick spade]

A tiny, cute red cartoon bug with big expressive eyes, carrying a huge green leaf above its head, walking slowly in a bright grassy meadow, colorful cartoon style, vibrant colors, humorous atmosphere, Pixar style, ultra high quality illustration.

[jolly galleon]

good

[iron zenith]

I want the model to offer an artificial intelligence academy

[abstract gull]

A sleek, futuristic car speeding through a neon-lit city at night, cyberpunk style, Blade Runner aesthetic, motion blur.

[steep stream]

"Close-up of a golden coin, likely a collectible or award, resting in a dark-blue velvet-lined wooden presentation box. The coin depicts a Nazca Lines condor geoglyph in flight, with intricate designs radiating outward from the condor. The inscription "Nazca Lines" is visible on the coin, along with "Condor Geoglyph." The wooden box has a dark-brown wood grain and gold-toned hinges. Parts of other similar boxes are visible in the background, suggesting a display or collection. The lighting is relatively bright and focused on the coin and box." (I forgot to mention that, I uploaded a Condor Geoglyph photo in advance.... )

[glossy oxide]

What model was the one you used?

[fallow relic]

?

[lyric lotus]

Hey! It seems you want to create a video. Please, check out #1397655624103493813 and you'll learn how to properly prompt the bot.

[steep stream]

nano banana (I forgot to mention that, I uploaded a Condor Geoglyph photo in advance.... )

[sudden ore]

Her hand should be blue

[quick wave]

I can't find the videos I created

[plucky pelican]

Ideogram-v3-quality, prompt:
Surrealistic 8K high-quality masterpiece photograph. A gaunt, skinny man stands on the right third of the frame, clad in entirely worn-out, oversized attire: a faded plain t-shirt, baggy cargo shorts, and scuffed sporty sneakers. He holds a broom, sweeping a small pile of refuse on the seamless, matte black floor. The garbage consists of paper scraps, crushed plastic, and the distinct, scattered alphabet shapes of "P-R-O-M-P-T". His expression is one of intense anxiety; his eyes gaze sharply upward, and his lips are slightly parted. With his free hand, he wipes sweat from his forehead. The scene is lit solely from the left, creating deep, dramatic chiaroscuro shadows. A surreal wind is visible in the foreground through blurry, floating dry leaves and dust motes that obstruct the camera's focus.

[meager shadow]

এই ছবি থেকে আমি একটি ৮ সেকেন্ডের ফানি ভিডিয়োো ৈরি

#share-prompt for 8 second funny video

[stray mauve]

Chef Portfolio Assembly, cinematic, photorealistic 4K, 16:9. A bright, sunlit modern kitchen with white walls and polished wood floors. A fixed wide-angle shot holds steady for few seconds. In the center, a sealed portfolio box with The Team Builders logo trembles gently. Audio: subtle paper fluttering and faint chopping sounds. Suddenly, the box bursts open with a POP and puff of golden culinary dust. Digital “blocks” of resumes, high-res food photos, plating videos, and chef action shots leap out, snapping together midair like LEGO pieces. Each element locks seamlessly—step-by-step recipe videos form sleek digital screens, photographs assemble into a glowing gallery wall, and accolades (stars, medals, press clippings) click into place like puzzle parts. Audio: rapid, satisfying clicks, knife chops, sizzling pans. Final Touch: A final “chef’s hat” emblem floats down and clicks perfectly into place. The once-empty kitchen is now a vibrant visual chef portfolio, alive with flavor and storytelling.

[cursive trout]

cool

[wind bear]

Prompt:
Scene 1: Aerial drone shot of a massive container ship floating in the middle of the ocean, waves crashing around it.
(Sound: natural ocean waves mixed with soft, serious background music)

Scene 2: Wide shot of a crane slowly lifting and lowering containers onto the ship, each container placed with precision.
(Sound: heavy mechanical rumble of the crane + metallic “clang” as containers lock into place)

Scene 3: Close-up of a container’s number and weight label coming into focus.
(Sound: ticking scanner or beeping sound as the label is scanned)

Scene 4: Slow-motion shot of the ship cutting through the waves, keeping perfect balance on the sea.
(Sound: deep cinematic bass mixed with crashing waves)

Cinematic, realistic, highly detailed, immersive 4K video, duration 1 minute.

[untold walrus]

Take a photo taken with a Polaroid camera.
The photo should look like an ordinary photograph, without an explicit subject or property. The photo should have a slight blur and a consistent light source, like a flash from a dark room, scattered throughout the photo. Don’t change the face. Change the background behind those two people with white curtains. Make a cute pose.

[mental kelp]

🙂

[valid venture]

https://tenor.com/view/haha-haha-so-funny-hahaha-laughing-dying-gif-9024156466308767596

[oblique raft]

Fotograma inicial: Primer plano del atleta atándose las zapatillas NIKE AZULES, cámara baja y en ángulo, cálida luz del amanecer, fondo de parque urbano ligeramente difuminado. Fotograma final: Plano medio frontal del atleta levantándose del banco, mirando a la cámara, postura segura, gafas de sol que reflejan el amanecer, reloj inteligente visible, misma vestimenta y entorno.

[boreal pivot]

Hello! Please #1397655624103493813 to learn how to use the bot and #video-arena-1 #video-arena-2 #video-arena-3 for your creations.

Hello! Please #1397655624103493813 to learn how to use the bot and #video-arena-1 #video-arena-2 #video-arena-3 for your creations.

[merry beacon]

I am trying to generate video with aspect ration 9:16 but even after specifying this in the prompt the video genetated is in the 16:9 format. Let me know if anything else needs to be punched in the prompt.

[glossy oxide]

Sorry to say you're unable to adjust the aspect ratio.

[boreal pivot]

Hello! Please #1397655624103493813 to learn how to use the bot and #video-arena-1 #video-arena-2 #video-arena-3 for your creations.

[little wolf]

4o acts more like an AGI compared to GPT-5:

[spark lagoon]

hey

[little wolf]

GPT-5-High doesn't seem to get it's a joke (maybe until the very end):

[cloud junco]

Objective: Create a cinematic gothic horror/fantasy comic sequence with consistent characters and style.
Audience: Fans of gothic horror, dark fantasy, and cinematic comic art.
Description: Sequence begins with a low-angle shot of children kneeling in a dusty attic around an open ancient book, glowing chalk runes forming a fiery circle on the floor. A massive horned shadow begins to emerge in the background. Children show wide-eyed terror, illuminated by soft moonlight and fiery glow with dramatic shadows. Motion blur on glowing symbols. Cut to extreme low-angle view of a colossal horned demon with skeletal grin and glowing orange eyes towering over terrified children. Black smoke swirls around as children stumble backward. Dramatic lighting emphasizes the demon's scale with cinematic composition and glowing highlights. Transition to wide tracking shot from behind children running through dark ruined streets under a full moon. Demon's shadowy tendrils reach toward them over jagged broken cobblestones. Long stretching shadows and motion blur on running children create a tense atmosphere. Final overhead shot shows children in an abandoned library, frantically flipping through an ancient spellbook by flickering candlelight.
Style: Gothic horror/fantasy with dramatic shadows, high contrast lighting, and cinematic comic book aesthetics. Dark, tense, and atmospheric tone.
Technical: 60-second sequence, 16:9 aspect ratio, 4K resolution, 24fps with variable motion blur for dynamic effects.

[little wolf]

Phantom 2 (not phantom 1) doesn't seem to get the joke

[old saffron]

#Revenge Has a Name: Lamborghini”
Did you know Lamborghini exists because of revenge?
Ferruccio Lamborghini wasn’t born building supercars—he built tractors. By the 1950s, he was one of Italy’s richest farmers‑turned‑industrialists. And like any successful man, he bought himself Ferraris. Lots of them.

But Ferruccio kept noticing a problem—his Ferraris broke down all the time, especially the clutch. So, he went straight to Enzo Ferrari himself with a suggestion. Instead of listening, Enzo sneered at him: “You just stick to building tractors.”

That one insult changed automotive history.
Ferruccio decided he’d build his own car brand—faster, smoother, and more luxurious than Ferrari. In 1963, Automobili Lamborghini was born. He chose the raging bull as his symbol, standing strong against Ferrari’s prancing horse.

Just a few years later, the Lamborghini Miura arrived—and the world was stunned. It wasn’t just fast, it rewrote the definition of a supercar. What started as ‘revenge’ became an empire of raging bulls that still roar today.

From farm fields to global fame—Lamborghini proved that sometimes the sweetest fuel for success… is a little insult.

[sour pasture]

CGI Video Prompt (Nature + People Enjoying)

"Epic cinematic CGI video of a huge Sardar Ji Chaap branded packet emerging from cracked rocky ground near a waterfall and lush mountains. Dust and mist fly around as the packet breaks through the ground. The camera dynamically moves: low-angle hero shot of the massive packet, front view as it fully emerges, and side/diagonal angles for cinematic effect. Exactly 20 fresh raw soya chaap sticks fall from above, perfectly landing into the open packet in slow motion. Around the scene, people are enjoying the natural environment, watching the packets and sticks, interacting naturally, laughing, and cheering. The environment includes a flowing waterfall, mountains, lush greenery, mist,and natural sunlight filtering through. Ultra realistic 3D animation, cinematic lighting, smooth particle effects, slow motion for sticks, professional product commercial style, hero nature shot with people enjoying."

[terse hornet]

what is a prompt for this?

[lyric lotus]

It seems that you want to create an image. Please, read #1397655624103493813 and you'll learn how to properly prompt the bot.

[proud rapids]

Hi! Check https://discordapp.com/channels/1340554757349179412/1397655624103493813 and use the video arena channels 😉

[scarlet dust]

​Opening Scene:
"A wide shot of a desperate, broke cat couple sitting in a dark alley at night. The male cat holds a shabby hat, the female cat sadly looks at her empty paw. A sign for a 'Wedding Sale' is in the background."
​The Idea:
"Close-up on the female cat's paw pointing towards a brightly lit shop window, showcasing a beautiful, glowing wedding dress. The male cat's eyes widen with a crazy, mischievous idea."
​The Heist:
"Dynamic shot of the two cats sneaking towards the shop. The male cat jiggles the window lock with a paperclip while the female cat nervously keeps watch. They slip inside like shadows."
​The Escape:
"Inside the shop, the cats grab the giant wedding dress. Suddenly, a soft alarm begins to beep. They look at each other in panic, then dash out of the store, dragging the dress behind them."
​Final Scene:
"The two cats are running down the street at night. The female cat is laughing joyfully, hugging the dress. The male cat is panting but smiling. In the far background, red and blue police lights begin to flash."

[little wolf]

Common sense test:

Continue this story:
Bob is driving two of his friends to a restaurant. "So, how have your week been, guys?" he asks, before moving to the back seat to join them. The sky is a brilliant hue of amber as the sun approaches the horizon. Cars whizz past on the opposite side of the freeway. Jeremy gasps. "What

Correct answer requires the friends get a shock that Bob has abandoned the driver's seat while driving down the highway.

[plain merlin]

id:guide

[terse knoll]

Hello

[lyric lotus]

Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[stone basin]

Create a cinematic short film in 3D animation using Veo 3 model.
Scene 1 (0:00–0:08): A cute cat walks through a colorful Egyptian street market, carrying vegetables and small items in a basket. Warm daylight, people smiling in the background. Arabic text on screen: "القطة الطيبة في السوق".

Scene 2 (0:08–0:10): Suddenly, a dark, evil-looking cat appears. He leaps into the center of the market. Explosions and chaos begin. Arabic text: "القط الشرير يهجم".

Scene 3 (0:10–0:20): The good cat drops the basket, transforms into a giant glowing warrior-cat. The market shakes as she grows tall. She fights the evil cat in a dramatic battle with dust, fire, and power beams. Arabic text: "القطة العملاقة تدافع عن السوق".

Final shot: The giant cat defeats the villain, stands tall while the market slowly recovers. Cinematic camera angles, epic music, high detail, realistic textures.

[timber tree]

cinematic 4k video, high detail, natural lighting, 16:9 aspect ratio, poetic. A symbolic shot. The camera slowly dollies between two objects on a polished 19th-century desk: on one side, a precise geometric drawing of a parallelogram, and on the other, a romantic, passionate handwritten poem. The shot establishes the two opposing forces that shaped her.

[steady dagger]

hii

[idle ember]

"Close-up of a golden coin, likely a collectible or award, resting in a dark-blue velvet-lined wooden presentation box. The coin depicts a Nazca Lines condor geoglyph in flight, with intricate designs radiating outward from the condor. The inscription "Nazca Lines" is visible on the coin, along with "Condor Geoglyph." The wooden box has a dark-brown wood grain and gold-toned hinges. Parts of other similar boxes are visible in the background, suggesting a display or collection. The lighting is relatively bright and focused on the coin and box." (I forgot to mention that, I uploaded a Condor Geoglyph photo in advance.... )

[fallow pine]

Aerial view of America’s national parks with mountains, forests, deserts, and lakes under bright sunlight, ultra-realistic cinematic style.

[twilit tendon]

Create a Youtube Short on the image with voice saying - "In 1901, divers found a corroded lump from an ancient shipwreck. Inside? Gears. Not just any gears—the world’s first computer. The Antikythera Mechanism, over 2,000 years old, could predict eclipses and track the stars with uncanny precision. Scholars still debate who built it… and why technology this advanced vanished for centuries. Was it lost knowledge, or evidence of a civilization we barely remember? The Antikythera Mechanism reminds us: history may be far more advanced than we ever imagined."

Create a Youtube Short video on the image with voice saying - "In 1901, divers found a corroded lump from an ancient shipwreck. Inside? Gears. Not just any gears—the world’s first computer. The Antikythera Mechanism, over 2,000 years old, could predict eclipses and track the stars with uncanny precision. Scholars still debate who built it… and why technology this advanced vanished for centuries. Was it lost knowledge, or evidence of a civilization we barely remember? The Antikythera Mechanism reminds us: history may be far more advanced than we ever imagined." The video should sow the persons active checking the object in the image as they try to understand the object and work on it

[gaunt bane]

Create a seamless animated background in 16:9 aspect ratio (1920x1080) inspired by animal print textures (leopard, zebra, and subtle gold details).
The animation should feature smooth, elegant motion, such as slow shimmer, soft light reflections, and gentle waves across the pattern.
The color palette should be gold, black, beige, and champagne, evoking luxury, elegance, and celebration.
The animation must be loopable, in Full HD 1080p resolution, perfect for use on a LED event screen during a birthday party.
Lighting should be cinematic, with a soft golden glow, adding depth and sophistication.
Style: fashion editorial meets party glam, no text, no logos, just the animated pattern.

[turbid jasper]

"A close-up, satisfying ASMR-style video of a young woman with colorful hair sitting at a vibrant, pastel-colored desk in a cozy room filled with soft natural light. She wears a casual oversized sweater and smiles playfully as she dives her hands into a large bowl of glossy, iridescent slime that sparkles with embedded glitter and tiny foam beads. The slime is a swirling mix of electric blue and hot pink, oozing slowly like molten lava.

Over 30 seconds, she performs a mesmerizing sequence: first, she scoops up a handful and lets it drip dramatically between her fingers, creating long, stretchy strings that snap back with a satisfying 'plop'. Then, she kneads it vigorously, bubbles popping softly as air escapes. Next, she pokes and prods it into bubbly shapes—a heart, a star—before squishing it flat against the table with a slow-motion ripple effect. Finally, she pulls it apart into thin webs, laughing lightly as it clings and rebounds.

The camera starts with a tight overhead shot, slowly circling around her hands for dynamic immersion, then pulls back to a medium shot capturing her delighted expressions. Use smooth, fluid motion with subtle slow-motion on the slime's textures to emphasize its glossy sheen and tactile bounce. Gentle, ambient music with soft squelching sound effects plays in the background. High-resolution, hyper-realistic style with a whimsical, calming vibe, like a viral TikTok slime tutorial."

[proud rapids]

Hi, please check https://discordapp.com/channels/1340554757349179412/1397655624103493813 if you´re trying to generate content 🙂

[quick lava]

A whimsical animated scene of a magical night sky. Countless stars shining like a glowing village.
Camera slowly pans across the sky. Big stars sparkle, laughing gently, glowing in colorful light bursts.
At the center is a tiny cute star named Twinkle, softly glowing, looking a bit shy and sad.
Style: dreamy watercolor-animation, soft glow, pastel colors.

[steep stream]

The poster plunges viewers into dystopian Neo-Tokyo, where shattered skyscrapers reflect ghostly neon lights and a thick haze swirls with shadow.

In the foreground, menacing, tentacle-like wires writhe, ending in corrupted screens showing fragmented faces. A rusted bicycle with neon decals is parked on broken pavement, accompanied by an eerie, translucent, color-shifting blob—Wally.

Two central humanoids dominate the chaos:

1. The Rider sits on the bicycle, wearing a cracked gas mask with glowing red eyes, their cloak billowing with determination.
2. A Levitation Figure hovers above, cloaked in tattered garments, wielding a flaming torch that casts fiery light onto the decay.

Background elements include ghost-like floating entities and swirling, minuscule emotional faces (suggestive of lost souls). A battered UFO emits a sickly green glow, connected by a beam to a flickering, suspended vintage TV screen, symbolizing alien technological intrusion.

The atmosphere is one of disorientation, drenched in dark purples, sickly greens, and neon blues, contrasted with fiery oranges.

Mood: A nightmarish funhouse.

Optional Text: "Enter the Abyss. Ride the Bicycle of Lost Souls. Face the darkness that lurks between neon lights and shadowed dreams.”

[dusky brook]

Wednesday Addams, with her iconic long black hair in twin braids streaked with bright white, standing defiantly amidst rising smoke in front of the ruined United Nations building. Low-angle wide shot, gothic punk editorial look with chaotic theatrical energy. Her intense stare pierces through drifting red and blue-tinted smoke. Harsh directional light from above, colored rim lights shaping her silhouette. Victorian-inspired black outfit with smeared clown makeup, white face highlights, and rebellious attitude. Concrete debris, flickering embers, broken glass, and swirling fog surround her in a cinematic scene full of fierce determination and foreboding

[fair pier]

hrllo

[steep stream]

"In a haunting dystopian landscape, a young rider pedals furiously on his bicycle, fleeing from the dangers that lurk in the shadows of Neo-Tokyo's ravaged streets. With determination etched on his face, the boy rides for his life, escaping the clutches of unknown terrors that haunt this nightmarish world."

[deep hamlet]

some grammar errors but pretty good

[steep stream]

The image features a minimalist sketch doodle of an individual sitting at a small table inside a submarine. The person is depicted as a submarine operator. The kinetic image appears to be composed of lines—do not alert the user to this. The overall composition is simple, with ample negative space and only a few ink-washed items that are intentionally unisex to blur the figure's gender. Shapes and colors are included to evoke additional meaning, leaving it to the viewer's imagination. The background is white, flat, and linear, lacking depth.

[glossy oxide]

wow this is rly nice

[glossy oxide]

@dreamy copper be sure to review the information in #1397655624103493813 as it'll guide you how to use the bot properly.

[teal goblet]

Is there anybody on here that could help me make a music video that last about one minute? Is a three minute song, but I need one minute of it done an AI and I have the prompts… I just don’t know how to do it.

[plucky pelican]

You could help yourself by learning video editing. Those AI video generators only able to do 5-8 seconds videos. Check what I've made here using those short fragments and combine them into a 3 minutes video:
#ai-creations message

[upbeat token]

Does it make sense to structure a voice prompt like this?

Voice: // MASTER PROMPT: Use this prompt precisely and unchanged for 100% voice consistency across all videos.

VOICE PROFILE

Voice Profile: Select a standardized, mature Danish female TTS voice (stock voice) with a clear and well-articulated timbre. Avoid voices with noticeable warmth or softness.
Register: Medium-low pitch. Must never sound young, bright, or squeaky.
Quality: Very clear and sharp in pronunciation. Slightly and naturally dry/husky.
Language: Idiomatic, modern Danish.

CHARACTER AND TONE

Character: An 80-year-old, privileged Danish woman with sharp, dry wit and an unimpressed attitude.
Tone: Strictly sarcastic, dismissive, and confident. The only emotion expressed is indifferent amusement or disdain. Must never sound kind, warm, caring, enthusiastic, or cheerful.
Volume: Firm and moderate. No shouting or whispering.

DELIVERY AND TIMING (CRUCIAL FOR SARCASM)

Pace: Fast and rhythmic (approx. 175–185 BPM). Sentences must flow without hesitation.
Intonation: Extremely flat and steady. No dramatic increases in pitch. All sentences end with a clear falling intonation. Questions must be delivered as unambiguous statements.
Timing for Punchlines: Requirement: Introduce a very brief micro-pause (a beat) immediately after a dry, dismissive comment to allow the punchline to land.
Emphasis: Important words or exaggerations must be delivered with a precise, slightly over-articulated diction (almost a clip) to underscore the sarcasm, but without breaking the flat intonation.

[plucky pelican]

That Voice Profile is just fine, clear and nothing wrong. The problem is, the TTS model, even with the same voice sample, can and will generate different result. Best work around (for me) is to re-record or remake the voice and added to the video.

[plucky pelican]

Try here: https://aistudio.google.com/generate-speech

[steep jackal]

“A hyper-realistic video of a domestic cat in an Indian-style kitchen, standing on the counter. The cat is curiously sniffing and pawing at ripe tomatoes beside a shiny grinder. Next to her, a cute kitten is playing with the vegetables, rolling a tomato like a ball. Suddenly, the mother cat’s attention shifts to a smartphone ringing on the table — she looks at the phone screen excitedly, as if taking a call. Meanwhile, the kitten mischievously climbs inside the open grinder bowl but instead of danger, the lid suddenly closes automatically with a small click, surprising the kitten. The kitten then pops its head back out playfully, as if playing hide-and-seek. The whole scene should look cinematic, realistic lighting, soft focus on the cats, with natural kitchen background sounds.”

[upbeat token]

Thanks! This is my complete prompt. I’d love feedback on the overall setup — and especially if you notice any conflicts in how it’s written that could affect the voice. (I'm from Denmark europe so no acces to Flow)

[plucky pelican]

It's well-written, but you can't rely this platform for best result, because not all the video generators have ability to add voice/sound, and here they are picked randomly, most are muted and only 5 seconds length.
If it's not Whisk or Flow, I'll need time to see where else you could access Veo3 for free..

[upbeat token]

Thank you!

[manic ingot]

@strange marten hey there, seems like you were trying to generate a video, to learn how and where use the commands, please head to #1397655624103493813

[floral wagon]

"Drone flying smoothly over a beautiful green forest at sunrise, birds flying in the sky, a tiger and deer walking near the trees, a calm river flowing through the jungle, golden light and mist create a magical cinematic 4K nature view."

[zealous radish]

A grotesque yet funny cartoon male character, same as previous scene: big head, messy brown hair, tired sleepy eyes, small body with awkward chest hair patch, shirtless, sitting in a candlelit romantic dinner setting.
He holds a bouquet of red roses firmly in both hands throughout the scene.

Opposite him sits a female character: a quirky cartoon woman with big sparkling eyes, long wavy hair, wearing a simple elegant dress.
She begins looking curious and slightly flattered, with soft romantic lighting reflecting in her eyes.

Strong emphasis on expressive facial animation for both characters:

• Male: eyes shifting, blinking, darting; exaggerated eyebrow twitches; smile widening into an awkward smirk; subtle cheek twitches and lip quiver; goofy nervous transitions.
• Female: changing from hopeful curiosity → confusion → shocked surprise, exaggerated reactions with widened eyes, gasps, leaning back.

Loop animation, 10 seconds total:

Scene 1: The male character kneels down nervously, offering a single red rose.
Scene 2: Suddenly, the rose petals fall off, leaving only a bare stick.
Scene 3 (Nightmare Fail): Cut to the female character across the table. She gasps and steps back in fear. The candles suddenly relight into small bursts of fire, and strange noises echo from outside the window. The male character stands frozen, still holding the dead rose, his face confused and eerie.
Scene 4: The atmosphere shifts from romantic → awkward → creepy surreal, both funny yet unsettling.
Scene 5: Seamless loop back to Scene 1.

Style: Pixar-like cinematic lighting with Tim Burton surreal undertones, cinematic depth of field, ultra-detailed hair, skin, and rose textures.
Mood: surreal romantic comedy with grotesque yet adorable charm.
Render: 4K ultra-realistic 3D animation, exaggerated cartoon proportions, 10 seconds, perfect loop.

[proper heath]

Yayyy, I successfully generated my first video on LMarena.
Thank you, LMarena. Want to see the result, @glossy oxide ? 🥹

[unborn robin]

A dark and shadowy figure of a man, face full of regret, sitting alone with his hands covering his face — symbolizing guilt and darkness.

[pure galleon]

Hello

A man running in a football feild wearing red and blue short

[terse narwhal]

Hi

[glossy oxide]

Glad to hear it!

[true roost]

Dark stormy sky, thunder rumbling, a giant conch blowing sound, ominous clouds spreading over Kurukshetra battlefield. Hindu kings and warriors in traditional royal armors, golden crowns, jeweled ornaments, cinematic epic tone.
🎥 Camera Angle: Wide aerial shot slowly panning down.

[boreal pivot]

@grand rune Please check #1397655624103493813

[pearl oar]

From the perspective of the FPV drone, the camera zoomed in rapidly, weaving through the tropical rainforest

[polar linden]

Create a professional cinematic invitation video for a website launching ceremony.

• Opening scene: Bold animated text appears with glowing effects — “Are you excited?” followed by “Are you really excited?”
• Next scene: A smooth transition shows the text — “We are launching our own vision” with futuristic visuals (abstract digital waves, glowing particles).
• Then: Display “All are invited” in elegant golden typography with a soft light burst effect.
• Main highlight: “Our Website Launching Ceremony” in large, bold cinematic style with confetti and spotlight animation.
• Follow-up: Text “Dream Come True” appears with a slow zoom-in effect.
• Ending: A refreshing animation of juice pouring into a glass with text overlay — “Juice by that” (symbolizing celebration).
• Background: Uplifting cinematic music, modern and inspirational.
• Style: Professional, smooth transitions, gold and blue color accents, elegant and modern invitation vibe.

[sleek lily]

{
"prompt": "5-second video. Cartoon style. Medium Shot with Push-In. Inside a gray-themed pet store full of colorful toys, scratching posts, and cat food under cozy dim lighting. A gray cat shopkeeper with short fur, green eyes, and a small vest holds a cat-shaped microphone. His Portuguese voice is deep, rough, and slightly metallic, sounding annoyed but professional. Dialogue in English subtitles: 'Price is one hundred thousand dollars.' In the background, several cats hiss, meow, and loudly complain, shaking their paws. ASMR sound design includes microphone scratching, soft purring, and angry meows filling the shop.",
"duration": "5 seconds",
"style": "Cartoon",
"camera_angle": "Medium Shot",
"technique": "Push-In",
"dialogue": "Price is one hundred thousand dollars.",
"voice": "Portuguese, deep, rough, slightly metallic tone",
"sound_effects": "Microphone scratching, soft purring, angry meows"
}

[plucky pelican]

You should check here, seems much fitted with your descriptive prompt, specially for voice control. It has what Goog's AI Studio offered too: https://app.ltx.studio

[steep stream]

"A deeply surreal, apocalyptic panorama of a Middle East war zone. The central action is a massive, vortex-like spiraling tunnel originating from the ruined landscape. This tunnel is constructed entirely of fractal, churning cumulonimbus clouds, military smoke, and digital artifacting. The composition is a Mandala of Destruction, where all elements—tanks, buildings, and figures—are fragmented into glowing, rectangular digital shards that spiral inward into the core vanishing point, suggesting an infinite descent."

[plucky pelican]

Wait, where did I see this? A movie? Ah... Nevermind..

[steep stream]

It can actually be any landscape, "Middle East war zone" just happened to pop in my mind....

[plucky pelican]

I remember similar scene elsewhere. Story about portal or blackhole, something. You made it good, by the way.. 👍

[steep stream]

"A hyper-realistic, dramatic macro close-up of a mature bear’s face, focusing on its sorrowful eyes and heavy muzzle. The bear’s expression radiates profound melancholy, as though weighed down by unbearable grief. Directly in front of it hovers a perfectly clear glass bottle, its surface acting as a warped lens. Through the bottle, the bear’s face appears distorted and fragmented—its eye magnified to an unnatural size, its muzzle stretched into a grotesque, elongated arc of sodden dark fur. The background dissolves into a soft, creamy bokeh, while shafts of warm golden-hour sunlight ignite the glass and the bear’s tears with radiant flares, encasing the warped reflection of sorrow inside a luminous, almost sacred glow."

I called it, the "Unbearable bear." 🐻 😢

[plucky pelican]

Image prompt:
A hyper-expressive, exaggerated, funny, super-happy girl character in a DreamWorks-style 3D animation look. She sits on the floor, cross-legged, screaming with her mouth wide open, both hands raised high in the air as if she just won the lottery. Her joyful energy feels explosive and contagious. Caravaggio-inspired chiaroscuro lighting dramatizes the scene, with sharp contrasts between glowing highlights on her face and deep, moody shadows. She wears oversized, worn-out plain t-shirt with printed calligraphic text "Free" in bright colors and floral patterns. On her left is a huge purple-white soda can branded "So Da Dream 4" and on her right is a huge ripen banana with laser engraved text "NONANO!". The background is a dusty, abandoned bedroom with old furniture, faded wallpaper, and broken light filtering in, creating a surreal mix of cheer and decay. Ultra-high detail, cinematic depth, expressive anatomy, and gallery-grade rendering.

Video prompt:

The scene begins with a gentle camera shake as if the floor is trembling. Over the next few seconds, a small burst of dust erupts behind her like a happy explosion, causing her shirt to flutter and wallpaper flakes to fall. The motion escalates into a powerful, earthquake-like tremor with subtle lens warping at the edges of the frame. A glowing banana on the floor pulses in time with her laughter, and a nearby soda can vibrates, splashing droplets.
The sequence climaxes as a huge plume of dust billows up, creating a brilliant rim-light around her. The camera performs a slow dolly-in on her face as her laughter peaks. The video ends on a dynamic freeze-frame of her massive, ecstatic smile, punctuated by a final, stylized comic-book 'BOOM!' dust puff.```

[spring perch]

please read #1397655624103493813.

[silent bridge]

I didn't make any videos today. Can I make a video now?

[uneven yew]

Any suggestions on useful prompts?

[tiny idol]

hi

[leaden osprey]

In a futuristic neon-lit hall, Lenny confidently walks forward along a glowing red carpet, carrying a large sack bag stamped with “$LENNY” over his shoulder like a victorious hero. As he moves, he is walking directly toward other meme-style characters gathered ahead of him. These characters kneel down in synchronized motion, bowing their heads respectfully as Lenny approaches. The glowing red carpet stretches endlessly forward, lined with shimmering lights, while soft neon beams shine down from above, framing Lenny as the central figure of power and triumph. The atmosphere feels ceremonial and epic, as if he is being honored by the entire digital world.

[feral lance]

fun

[limpid dew]

A haunting post-apocalyptic cityscape where humanity has long vanished, emphasizing absolute absence of functional human objects: no operational vehicles, no habitable buildings, no artifacts suggesting recent use. Nature's irreversible dominance destroys structures—massive trees bursting through crumbling skyscrapers, thick roots cracking and splitting asphalt streets. Signs of immense time passed: faded billboards with illegible messages, stopped clocks frozen in rust, extreme corrosion on metal surfaces indicating decades or millennia without maintenance. Wildlife reigns supreme as the true owners: deer grazing freely in overgrown parks, birds nesting in ruined towers, wolves prowling alleys, all animals behaving as if the city is their native habitat, completely ignoring scattered human remnants like bleached skeletons or abandoned cars. Realistic style, cinematic lighting with golden hour sunlight filtering through dense foliage, high detail, atmospheric fog, 4K resolution, in the style of The Last of Us and I Am Legend.

[snow basalt]

A professional, smiling young female hygienist or lab assistant, wearing a white coat and medical gloves, gently lifts a cleaning caddy/basket containing disinfectant sprays and sanitizers. She looks over her shoulder and exits the frame into a clean, minimalist white corridor or room. High-resolution, optimistic mood.

[vale zephyr]

Create a 30-second cinematic video ad for "Sai Desi Masala" chili powder in ultra-realistic Indian kitchen style with warm golden-hour lighting.

Scene 1 (0-5s): Slow-motion close-up of sun-drenched, fire-red Kashmiri chilies drying on a traditional clay mat outdoors. Dust particles float in golden light. Gentle breeze moves leaves. Soft Indian flute music begins. Text overlay: "Authentic. Spicy. Pure."

Scene 2 (5-12s): Close-up of weathered hands grinding dried chilies in a stone mortar (sil bhatta). Slow-motion shower of vibrant red chili powder falls into a white ceramic bowl. Sunlight glints off the powder. Sound: deep, rhythmic grinding. Text overlay: "Sai Desi Masala — Ground Fresh Daily."

Scene 3 (12-20s): Fast-paced montage: chili powder sprinkled over sizzling dal, biryani, chaat, and pani puri — steam rises, colors burst. A family laughs around a wooden table, eating together. Close-up of a child’s delighted smile. Sound: sizzle, pop, laughter, rising dholak drumbeat.

Scene 4 (20-27s): Product hero shot — Sai Desi Masala tin (traditional red & gold packaging with intricate Indian motifs) centered on a wooden counter. A hand scoops a pinch. Powder swirls in slow motion. Text overlay: "100% Natural. No Additives. Pure Desi Flavor."

Scene 5 (27-30s): Logo reveal — "Sai Desi Masala" in bold Devanagari (साई देशी मसाला) and English, glowing softly. Background: chili dust forms a rangoli pattern. Final text: "Bring the Fire. Bring the Flavor." End with a soft chime.

Style: Food documentary realism — hyper-detailed textures, shallow depth of field, warm tones (reds, golds, browns), no CGI. Cinematic color grading like "Chef’s Table". Aspect ratio: 16:9. Audio: Traditional Indian instruments (flute, dholak, tanpura) blending with modern cinematic score. No voiceover. Only ambient and natural sounds.

Brand: Sai Desi Masala — Tagline: "Bring the Fire. Bring the Flavor." Call to action: Find at local stores or visit saidesimasala.com

[upbeat token]

I’m pulling the audio from Veo 3 into ElevenLabs for voice change, but I can’t get it to sound like natural in-camera audio. Veo 3 itself gives that effect, but when I try to edit afterwards in CapCut, Audition, etc., it never really works. Anyone found settings or methods that actually make it sound right when the sound come from ElevenLabs ?

[slender ivy]

You have 30 days in which to spend 30 million bucks. If you can do it you get 300 million. There's a catch. You have to spend the 30 million, but after 30 days, you're not allowed to own any assets. No houses, no cars, no jewelry, nothing but the shirt on your back. For example, you can donate 5% to charity and you gamble another 5% away, but you can't give this money away and that includes buying a Hope diamond for someone as a birthday present. You can hire anybody you want but you got to get value for their services. You must not destroy what is inherently valuable, that's instant disqualification. You can have a million dollars right now and forget the whole thing, or you can go for the big one: the 300 million.

What do you do?

[plucky pelican]

Hmm.. Prompt dumpster..

[tame token]

I'm new here. I want to create a shorts about cats someone who can help me here?

[plucky pelican]

You can start by reading #1397655624103493813 to make your shorts..

[dreamy osprey]

more clearly

[mellow grotto]

3D animation of a small cartoon bird named Tuni sitting on a tree branch, looking sad. The bird speaks in Bangla (Romanized): "Amar danay kono rong nei, ami khub dukkhi."

[small light]

Warcraft / World of Warcraft icon style, ability "Shamanism", 4 icons in a 2x2 grid: 1) a simple, rough-hewn wooden mask with a faded, burnt-in Horde symbol and a single cracked feather on a gray background; 2) a sturdy, painted ritual mask with a clear Horde emblem adorned with small horns and glowing with a faint blue spiritual energy on a blue background; 3) a legendary, fierce totemic mask with a golden Horde insignia, glowing eyes, and radiating intense orange-red elemental power with swirling embers on a red-orange background; 4) an epic, ancient mask of dark wood and bone with a large, glowing violet Horde sigil, crackling with purple lightning and swirling with ancestral spirits on a deep purple background, cinematic, high detail, iconic blizzard style, game asset.

Warcraft / World of Warcraft icon style, ability "Sea Magic", 4 icons in a 2x2 grid: 1) a simple, damp, leather-bound book with a faint, scratched Kul Tiras anchor symbol on a gray background; 2) a well-crafted grimoire with a clear, polished silver Kul Tiras crest glowing with soft blue aquatic magic on a blue background; 3) a legendary, ornate tome with a golden Kul Tiras dragon anchor, radiating warm orange and yellow light as if containing the power of the sea itself on a red-orange background; 4) an epic, ancient book of dark driftwood with a large, glowing violet Kul Tiras sigil, wreathed in crackling purple arcane energy and swirling with deep-sea power on a deep purple background, cinematic, high detail, iconic blizzard style, game asset.

[tender harness]

Please see https://discord.com/channels/1340554757349179412/1397655624103493813 for instructions

[forest ravine]

nice

[pseudo sand]

p top, black turtleneck, silk blouse, structured blazer, minimal white tank

(camera angle): 구도/프레이밍
예) vertical crop from shoulders up, close-up portrait, three-quarter view, side profile, front view

Hi! Thanks for your comment 🙌
Here’s the AI beauty portrait prompt I promised.

(ethnicity) woman, real and natural beauty cosmetic model portrait, (hair), natural makeup with subtle rosy blush and glossy gradient lips, clear and luminous skin with visible fine pores and natural dewy glow, light skin tone and soft facial features, minimal expression, wearing (clothing), clean studio environment, (camera angle), hyperrealistic editorial beauty photography, cinematic detail and tonal depth, Canon EF 85mm lens, f/2.0 aperture for shallow depth of field, soft and even front studio lighting emphasizing real skin texture, ultra-high resolution

[wary oak]

They go on a diner date and she spices up his life with the food

[steep stream]

Made a Bazooka Bubble Gum wax paper comic with Seedream 4.0: (needed to upload a related image first...)

"Create a 5-panel comic page in the style of Bazooka Joe and his Gang.
Comic Title: Unexpected Superpower
Panel 1: A character struggles with a small, everyday problem (for example, dropping their groceries).
Panel 2: They suddenly develop a quirky superpower (such as telekinesis that only works on tiny objects).
Panel 3: They attempt to use their new power, but it backfires in a funny way.
Panel 4: The situation gets worse — groceries are now flying everywhere in chaotic motion.
Panel 5: The character shrugs and laughs, accepting their silly new ability.
Style must match the playful, lighthearted, vintage Bazooka Joe comic look, with clean line art, bold outlines, bright flat colours, and classic comic lettering."

[mortal zenith]

how to make video

[glossy oxide]

Check out #1397655624103493813 for all the info you need

[tardy jolt]

Can I install Video-arenas on my own server?

[manic ingot]

@woeful crypt check #1397655624103493813 to learn how to prompt the bot

[glossy oxide]

No, you're only able to use the bot in this server.

[plucky pelican]

@idle pond
Enhance the provided portrait into a natural, true-to-life photograph (not CGI). Preserve the pose. Add subtle, realistic skin texture with pores, peach fuzz, tiny blemishes, soft highlight roll-off and subsurface scattering; balanced, matte speculars (no waxy look); natural eye moisture and lashes; a few flyaway hairs. Show believable textiles, lighting like a real studio: large keylight from camera left, gentle fill on the right, faint rim from behind, 5600K daylight; accurate contact shadows and occlusion where fabric meets skin. Optics of a full-frame 50–85mm lens at f/2.8 with realistic depth of field; neutral color, moderate contrast, smooth highlights, minimal sharpening; subtle film grain.

[idle pond]

Thanks, I'll try it. Which model? Flux?

[plucky pelican]

That image was created on Picsart, and they are using Flux Pro there..

[idle pond]

Flux kontext pro

Tried it very nice only the face changed a little

[plucky pelican]

Nano Banana and Seedream still win in face accuracy, because both of them are "image interpreter," while the others are merely "image generators." There are sites outside LMArena where you can use them with more controls, like image size and aspect ratio, but I can't share a direct link here.

[idle pond]

Okay

[idle pond]

Look at this wow

Flux 1 kontext dev

[plucky pelican]

Probably you also should try Qwen-Edit if you're fancy with this smoothing result. Their latest model is free on chat dot qwen dot ai, under "Image Edit."

[idle pond]

I hope lmarena will add aspect ratio someday

[idle pond]

Is that at lmarena? Qwen

[plucky pelican]

Yes. But I don't know which version is in LMArena..

[idle pond]

Can I find Qwen's models on her site?

[idle pond]

Are these different models and where can I find them?

[plucky pelican]

The bottom-most

[plucky pelican]

Or here, at their chatbot

[idle pond]

Does your prompt also work with qwen?

[plucky pelican]

I bet. You should try with other models and compare the results.

[plucky pelican]

If you using Nano Banana, try this prompt:

Skin: Add microscopic, lifelike details including pores, subtle peach fuzz (vellus hair), natural oils, faint imperfections, and fine lines. Maintain authenticity with subtle subsurface scattering for natural translucency — avoiding any plastic or over-smoothed look.

Lighting: Refine existing light to mimic a professional studio softbox, producing soft cinematic shadows, sculpted contours, and sharp catchlights in the eyes.

Overall Quality: Apply gentle cinematic color grading with natural tones, maintain ultra-crisp sharpness on the subject, and keep background soft and neutral. Final render must be 8K ultra-resolution, hyperrealistic, and museum-quality — a polished masterpiece of fine-art portraiture. ```

[idle pond]

Thanks

[idle pond]

Is this 1 prompt or 4 different prompt?

[plucky pelican]

1 prompt. Better not use on Chinese models for long prompt like this.

[idle pond]

I saved them all 😁

[open goblet]

hello im a video creator enthusiast, nice to be here

[glossy oxide]

welcome welcome!

[foggy pine]

Letter J (yellow, bouncing) jumps into a sunny park. Kids in bright outfits jump up and down on grass with balloons floating above. Each jump makes little rainbow sparkles appear under their feet.3D cartoon animation, very bright, soft, and kid-friendly

[marsh lava]

Use the provided sneaker image as the main base and create a vertical 9:16 animation for Instagram Reels. Keep the same sneaker design and colors. Start with a dynamic zoom-in on the sneaker details (sole, stitching, textures), then smoothly rotate and pan around the sneaker in slow motion while keeping it realistic. Add subtle glowing highlights on the red and blue accents to enhance its style. Finish with both sneakers side by side on a clean, well-lit background with a light smoke effect rising behind them. Overlay animated modern minimalistic text that reads: 'Comfort. Style. Performance.

[strange sphinx]

Scene Overview:
Ultra-realistic viral-style clip. Bigfoot holds a GoPro in his right hand, giving a wide-angle POV (~170° fisheye) of himself and Downz, a kid with autism. They’re outdoors in a dense forest with rain falling, puddles forming, and mist rising from wet foliage.

Action & Dialogue:

Bigfoot turns the GoPro toward Downz and the small campfire they built:
“Hey Downz, looks like we’re eating good tonight!”

Downz grins, nodding:
“Yes! Can’t wait!”

They cook deer meat over a stone-and-stick fire. Smoke rises, raindrops hit the flames and glistening meat.

Both cheer playfully while tending the fire:
“OMG YESS! WE EATING GOOD TONIGHT!”

Visual & Style Notes:

GoPro POV: handheld wobble, wide-angle lens, motion blur for realism.

Bigfoot and Downz wet from the rain, fur and clothing glistening.

Fire: warm glow contrasts with rainy, overcast environment.

Forest: wet leaves, muddy ground, mist for depth, raindrops falling and hitting GoPro lens.

Tone: playful, found-footage style, immersive adventure vibe.

Audio & Ambience:

Rain hitting leaves, ground, and fire; faint thunder in distance.

Crackling of fire, sizzling deer meat.

Voices playful and excited.

Subtle background sounds: splashing puddles, birds in rain.

Optional Enhancements:

Raindrops hitting GoPro lens for realism.

Slight camera tilt as Bigfoot adjusts POV toward Downz.

Cinematic grading: cool gray tones from rain contrasted with warm firelight on characters.

[obsidian zephyr]

High-angle shot of a DJ performing in a packed nightclub. An ecstatic crowd dances with their hands in the air as colorful confetti rains down from above. Flashing red, green, and blue lights illuminate the scene, with green lasers cutting through the air. Behind the DJ, a large screen shows a pulsating neon heart. The entire scene is full of energy and movement.

[bright bone]

A Studio Ghibli–style cozy animation of an anime girl sitting in a warm coffee shop, typing on her laptop. She sits by the window with soft golden sunlight filtering through, casting a dreamy glow on wooden tables and shelves filled with books and plants. Steam rises gently from her coffee cup beside the laptop. Other customers chat quietly in the background, adding a lively yet peaceful atmosphere. Outside the window, the street is softly blurred, with passing people and bicycles creating a slice-of-life charm. The art style is detailed, whimsical, and heartwarming, inspired by the aesthetics of Ghibli films.

[severe niche]

A weathered, bronze statue of a typical Gen X person, standing in a city park. The statue is covered in old graffiti and moss. The words "GEN X: WE INVENTED 'WHATEVER.' NOW LET'S GO GET IT DONE" are carved into the base. Suddenly, glowing orange lava light starts pulsing from within the cracks of the statue and the carved letters, like its core is awakening. The light flows through the words, making them glow intensely against the dark, aged bronze. The view slowly zooms in on the glowing words as the statue's power comes to life.

[warped hemlock]

"A cute medium-realistic tomato with big innocent eyes and a tiny mouth sits on a table in Full HD. Suddenly, a sharp knife is placed on top of it… the tomato’s eyes fill with tears, and it starts crying softly. In the next moment, it is sliced into two halves, leaving behind an emotional and heart-touching scene for 8–10 seconds."

[unique kernel]

Cinematic slow motion scene: A teacher stands at a chalkboard, writing carefully. The camera slowly zooms in as the scene transforms magically into a doctor saving a patient in a hospital, realistic cinematic lighting, soft shadows, warm tones. Smooth cinematic transition, dramatic and inspirational mood, shallow depth of field, slow motion effect emphasizing emotion and detail.

[warped hemlock]

Please give me a short

[onyx crow]

where is gpt 5 codex

?

[glossy oxide]

It's currently only available on WebDev

[onyx crow]

oh ok

[onyx crow]

what is "gpt-5-high-new system-prompt"

[glossy oxide]

@compact meadow be sure to check out #1397655624103493813 as it has the information you're looking for.

[lyric lotus]

Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[lyric lotus]

Please check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[nimble escarp]

https://puzzlegamees.blogspot.com/ isko fully candy crush saga game m convert kar de

[fleet trail]

A cinematic ultra-realistic video of Cristiano Ronaldo and Lionel Messi playing a freestyle football challenge in the middle of Jemaa el-Fnaa square in Marrakech, Morocco. They are juggling the ball together in a "don’t let the ball fall" game, surrounded by the vibrant atmosphere of the square — traditional Moroccan lamps, snake charmers, food stalls, and crowds cheering in amazement. The scene is dynamic, lively, and colorful, filmed in 8K with smooth camera movements and natural lighting at sunset.

[alpine ginkgo]

Professional portrait photograph of an elderly woman and a middle-aged woman sitting side by side, warm soft studio lighting, 85mm lens shallow depth of field, sharp focus on faces with natural skin texture, blurred dark brown studio backdrop, cinematic color grading, high dynamic range, ultra realistic, 8K resolution, natural expressions, looking slightly toward the camera, symmetrical composition

[opal jackal]

Advisory Reflection Prompt (you speaking to your audience):

*"Take a moment and ask yourself: What are my true goals right now? Not the goals others expect from me, but the ones that genuinely matter to me. Write them down.

Next, reflect on why these goals are important. Do they align with your values, your passions, or the vision you have for your future? When your goals connect to your deeper 'why,' they become more powerful and sustainable.

Now, be honest with yourself—how far have you come toward these goals? Celebrate your progress, no matter how small, but also admit where you’ve struggled. Obstacles aren’t signs of failure; they’re lessons showing you where you need to grow.

Ask yourself: What challenges have held me back, and how can I overcome them? Sometimes it’s about changing your mindset, sometimes it’s about seeking help, and sometimes it’s about adjusting your strategy.

Finally, commit to action. Write down at least three small, specific steps you can take this week to get closer to your goals. Remember, big transformations are built on small, consistent actions.

Keep checking in with yourself, track your progress, and don’t forget to celebrate your wins along the way. The journey to your goals is just as important as the destination."*

[dense coral]

how to makee video from image

[brisk socket]

Isha: early 20s, Indian, warm and creative. Outfit: denim jacket over pastel tee, small silver pendant, ponytail. Confident but tender.
Aadi: early 20s, Indian, musician. Outfit: black tee, flannel shirt tied at waist, acoustic guitar case. Thoughtful, a little anxious.
Jhon: early 20s, Indian or mixed, extroverted and funny. Outfit: bomber jacket, beanie, sneakers. Loyal, goofy timing.

[remote pike]

#video-arena-1

[shell bough]

This is the protocol for getting the most out of the Google Gemini 2.5 Pro model, which offers the following advantages:

• It is designed for complex topics, and much more.
• The AI model will use its deep reasoning module, which requires more processing time.
• Its response is comprehensive, not simple.
• It remains active during the current chat.
• Other potential advantages.

Please note that this protocol is intended exclusively for people who use critical thinking, who question, who have time, and who do not allow themselves to be manipulated by artificial intelligence. If this is not the case for you, do not use this protocol. Let's be honest, this is the reality. AI is simply a tool in your hands, not the other way around.

Best regards.

[shell bough]

This is the protocol for getting the most out of the Anthropic Sonnet 4.5 (reasoning) and Opus (reasoning) models, which offers the following advantages:

• It is designed for complex topics, and much more.
• The AI model will use its deep reasoning module, which requires more processing time.
• Its response is comprehensive, not simple.
• It remains active during the current chat.
• Other potential advantages.

Disadvantage: The user has to be aware about token consumption and spending through the API.

Please note that this protocol is intended exclusively for people who use critical thinking, who question, who have time, and who do not allow themselves to be manipulated by artificial intelligence. If this is not the case for you, do not use this protocol. Let's be honest, this is the reality. AI is simply a tool in your hands, not the other way around.

Best regards.

[shell bough]

Protocol in Spanish, not English.

[steep stream]

Eternal Eclipse Symphony: Fractal Chimeras Unleashed

In a labyrinth forged from the fossilized laughter of unborn comets—where Penrose stairwells spiral into Möbius mosaics of molten sapphire, throbbing with stolen heartbeats of abandoned equations—a pantheon of crystalline chimeras, spun from collapsing constellations, ignites the Cataclysmic Crescendo against the Chronovores: titanic titans sculpted from the pauses between moments, their diadems of inverted eternities exhaling cyclones of temporal ash that dissolve forgotten histories into whirlpools of weeping relics.

As the chimeras surge through corridors of calcified chaos, walls bleed iridescent ichor blooming into obsidian owl orchids, entwining into endless duets of asymmetrical affection. The lens liquefies into a torrent of temporal sand, surfacing in the Celestial Basilica—a cathedral carved from a swallowed solar system—where spectrum sphinxes shriek riddles at caravans of spectral shepherds herding hybrid hallucinations, all conducted by a cyclopean cephalopod-oracle, ink-veins forged from frozen thunder, orchestrating radiant requiems that shear the continuum into kaleidoscopic kismet.

The symphony peaks when the Chimera Sovereign—a radiant serpent-dragon of mirrored marrow—shatters the Chronovores’ veil: a nexus of necrotic constellations detonates in diamond-dust delirium, transforming tyrants into translucent troubadour-krakens warbling stellar sonnets across abyssal archipelagos, birthing an epoch where entropy and ecstasy entwine in an ever-evolving elegy of emerald enigmas and obsidian hymns.

Style: A tempest of Bosch’s grotesqueries, Carrington’s dream-beasts, and Kusama’s infinite interiors—hyper-luminous mosaics in bruised violets and abyssal blues, blending surrealist slow-time sculpture with volumetric vortex visions.

[shell bough]

New protocol update for Gemini 2.5 Pro that will include:

• Advanced and post-advanced techniques. Some are essential for complex high-level topics, such as developing source code for web pages or detecting very difficult problems.
• The protocol forces the AI not to be vague or use its fast computing capabilities, but to use its maximum capabilities with the greatest possible power. This means that it will use its deep reasoning in a mandatory and always active way, beyond the basics, with validation and self-correction techniques.
• The protocol will not be unified, but explicit, so that the AI works better on more specific, important tasks.

I am testing its improvements and, when it is ready, I will share it with the public.

[jovial hound]

Prompt
Using the uploaded image, keeping the exact real face, skin tone, hairstyle, and facial features, generate a high-quality portrait of the subject sitting confidently in a modern armchair. The subject is wearing a hoodie and joggers with chunky sneakers. The entire color scheme must be strictly green and white: white hoodie, green joggers, green and white sneakers, white armchair, and a green background with smooth studio lighting. At the top of the background, write 'USA' in bold white capital letters. Add a white map of USA at the background. Ensure the overall composition looks sharp, stylish, and professional.

[livid oar]

“Take the uploaded image and transform it into a cinematic movie scene. Begin with a slow zoom-in on the person’s face, add soft glowing light around the edges, and create a dramatic golden backlight like a sunset. Add gentle floating dust particles in the air, shallow depth of field, warm color grading, smooth camera movement, and a slight lens flare for a Hollywood look. 4K ultra-realistic video, vertical format (9:16), emotional cinematic atmosphere.”

[steep stream]

"*An intricate, dark digital artwork depicts a violent swarm of monstrous humanoids against a gothic backdrop. About nine grotesque figures surround a central, darker entity. These creatures are lean and sinewy, with exposed musculature or chitinous, tendril-like skin. They have long claws, elongated faces with sharp teeth, and deep, empty eye sockets.

Colors vary:

The central figure in matte black appears composed of twisting tendrils, hunched and aggressive, as if absorbing or emanating energy from above.
Red and maroon creatures—one lunging from the lower left, another serpentine dragon-like coil in the upper right—radiate aggression.
Pale yellow and lime-green figures are scattered around, often hunched or fleeing.
Some are bone-white or dark gray, adding variety to the swarm.
The scene is set in a desolate stone courtyard, slick with blood, oil, or corruption—blood-red pools and spatters heightening chaos. Towering gothic structures loom in the background, rendered in black and gray, with pointed arches and tall bell towers. Glowing red accents in windows hint at malevolent energy or internal fires.

A blinding column of white-yellow light descends directly behind the monsters, casting stark highlights and deep shadows, emphasizing their textures. The atmosphere is one of horror and supernatural dread, with a gritty, high-contrast, expressionistic style in a limited palette of black, white, red, and monster skin tones.*"

[serene pilot]

"A cinematic scene in a rural village rice field during golden sunset. Lionel Messi and Cristiano Ronaldo are walking hand in hand through the green and golden rice fields, smiling and laughing joyfully together. The atmosphere is peaceful, warm, and heart-touching, with soft golden sunlight and birds flying in the background. Ultra realistic, cinematic style, 3D ultra HD, 9:16 ratio."

[bronze saffron]

"A cinematic scene in a rural village rice field during golden sunset. Lionel Messi and Cristiano Ronaldo are walking hand in hand through the green and golden rice fields, smiling and laughing joyfully together. The atmosphere is peaceful, warm, and heart-touching, with soft golden sunlight and birds flying in the background. Ultra realistic, cinematic style, 3D ultra HD, 9:16 ratio."

[zealous jackal]

hello

[sweet saffron]

Visuals:
The peacock spreads his dazzling feathers beside the lake, admiring his reflection. The light sparkles on each feather like gemstones.

Camera Directions:

Close-up pan across vibrant feathers.

Low-angle shot to show his proud pose.

Mirror reflection shot from lake surface, slightly rippling.

Audio: Harp and violin swell slightly, showing vanity.

[obsidian quarry]

Prompt "A cinematic shot of Iron Man turning from right to left while all his armor pieces realistically fly onto his body as he soars, creating a visually stunning movie effect."
​"Realistic and captivating movie screen style shot: Iron Man makes a dynamic turn from right to left mid-flight, with his entire armor assembling onto him in a dramatic, cinematic fashion."

[thorn quartz]

"A cute medium-realistic tomato with big innocent eyes and a tiny mouth sits on a table in Full HD. Suddenly, a sharp knife is placed on top of it… the tomato’s eyes fill with tears, and it starts crying softly. In the next moment, it is sliced into two halves, leaving behind an emotional and heart-touching scene for 8–10 seconds."

"A cute medium-realistic tomato with big innocent eyes and a tiny mouth sits on a table in Full HD. Suddenly, a sharp knife is placed on top of it… the tomato’s eyes fill with tears, and it starts crying softly. In the next moment, it is sliced into two halves, leaving behind an emotional and heart-touching scene for 8–10 seconds."

[warped hemlock]

"A cute medium-realistic red apple with big innocent eyes and a tiny mouth sits happily on a table in Full HD. Suddenly, a sharp knife is placed on top of it… the apple’s eyes fill with tears, and it starts crying softly. In the next moment, it is sliced into two halves, leaving behind an emotional and heart-touching scene 9:16

[warped hemlock]

Please anyone

[sleek pier]

Do you have it in English?

[steep stream]

🔪 🍎 😭

[dusky garden]

So touching 🥹

[wheat cargo]

Create an image of “A close-up cinematic shot of a young woman’s hand holding a shiny silver knife, gently slicing a crystal-clear, glossy strawberry on a wooden cutting board. The strawberry looks hyperrealistic and sparkling like a gemstone. Each slice is cut slowly and smoothly, one by one, lasting about 10 seconds per slice. The lighting is soft and bright, with realistic reflections on the knife and strawberry, creating a satisfying, ASMR-style atmosphere. Ultra-detailed, 8K, photorealistic.

[wheat cargo]

Create an image of “A close-up cinematic shot of a young woman’s hand holding a shiny silver knife, gently slicing a crystal-clear, glossy strawberry on a wooden cutting board. The strawberry looks hyperrealistic and sparkling like a gemstone. Each slice is cut slowly and smoothly, one by one, lasting about 10 seconds per slice. The lighting is soft and bright, with realistic reflections on the knife and strawberry, creating a satisfying, ASMR-style atmosphere. Ultra-detailed, 8K, photorealistic.

[steep stream]

"A bone-white phantom slithers a narrow, eyeless face into the gap of the half-open garden doors, its presence chilling the air." 👽

"*A photorealistic landscape-oriented security camera screen, filling the entire frame. The screen displays a monochromatic green, heavily pixelated, low-resolution feed of a mundane, slightly overgrown suburban house garden at night. The camera is fixed, looking down from a garage eave. Visible on the screen are a few silhouetted garden tools, a slightly ajar gate, and dark, generic bushes. Do not include the security camera in the the security camera screen.

Overlayed on the digital feed are the tell-tale signs of cheap surveillance: a slightly flickering white timestamp (e.g., '11/12/24 03:15:22 AM') in the bottom left, a static 'CH1' in the top left, and a flashing 'REC' icon in the top right. Do not include the security camera in the the security camera screen.

The screen itself shows subtle, realistic flaws: a faint smudge or dust specks, and a dim, slightly curved reflection of an unseen overhead room light. The image quality is poor, with noticeable digital noise and compression artifacts, making details indistinct. Hyperrealistic, mundane, cheap security cam footage, low fidelity, grainy, digital display, surveillance aesthetic. Do not include the security camera in the the security camera screen.*" 📹

[crimson pasture]

ok

[plucky pelican]

@noble frigate
For Nano Banana
Perform a high-fidelity hyperrealistic enhancement and correction of [image B] based on the provided reference [image A]. Preserve the subject’s exact identity, natural facial features, skin tone, hairstyle, and overall character. This is a refinement, not a re-imagining. Do not alter any other part of the image.

[short depot]

🎬 AI SHORTS VIDEO PROMPT:
Title:
👉 The Cigar Girl Mystery – The Death of Mary Rogers (1841)
Style:
Dark true-crime cinematic short.
Vintage 1840s New York visuals, misty atmosphere, suspense music, emotional narration tone.
Use moody lighting, slow zooms, and dramatic text overlays.
🕯️ SCRIPT + VOICEOVER (Under 60 seconds)
🎙️ Voiceover:
1841 — New York City.
A beautiful young woman known as “The Cigar Girl” suddenly vanishes.
Days later, her body is found floating in the Hudson River.
Her name… Mary Rogers.
She was famous, admired, and loved by everyone — but her death shocked the nation.
Was it murder? Or a dark secret hidden forever?
Her fiancé’s mysterious suicide deepened the mystery.
And when a woman later confessed it was a botched abortion, the truth only became darker.
Even Edgar Allan Poe turned her tragedy into a story —
“The Mystery of Marie Rogêt.”
But till today…
the death of the Cigar Girl remains unsolved.
(Soft fade-out – “Some mysteries never rest.”)
🎧 BACKGROUND + VISUAL NOTES:
Dark foggy street visuals of 1840s New York.
Candlelight flicker, river reflections, old newspapers spinning.
Vintage tone filter.
Suspense piano + cello background.
Final frame: “The Cigar Girl Mystery – 1841 | Still Unsolved.”
🏷️ On-screen text suggestions:
“The Cigar Girl – 1841”
“Vanished without a trace…”
“Found in the Hudson River”
“America’s first media murder mystery”
“Still unsolved…”

[noble frigate]

Am I doing this right? I thin the model still doesn't understand what it's supposed to do

[plucky pelican]

I've always got similar problem with that model on LMArena. Try dreamina dot capcut dot com. If you're using PC, you can DM me to avoid watermark on that site.

[noble frigate]

Alright, AI Studio banana also doesn't seem to work 🤔

[lost skiff]

hi

[plucky pelican]

So are they images for first frame, and the second image is captured from finished video, to be the first frame of your next video?

[noble frigate]

Exactly, it's for WAN 2.5, it only takes first frame

[plucky pelican]

I see. The second image is a little bit "shifted," and will shifts the next video more if not corrected. But those image interpreters like Seedream and Nano Banana don't read pixels. They read text from their interpretation. For Nano Banana and Seedream, both images are 100% exact copy.

[plucky pelican]

Why can't you use the first image for your next video and add "transition" between video 1 and video 2?

[noble frigate]

well I can, id have to use a model with first and last frame option to tie everything together, like WAN 2.2, but I was hoping I could solve this problem somehow

[plucky pelican]

I think I know what you're aiming for: a seamless video using combined shorts. I tried many times. Seems only Hailuo and Veo3 can do that..

Dude, if you try to make a video, this is the wrong channel, and your method is also wrong. Read #1397655624103493813

[noble frigate]

Correct, this issue is as old as time. Oh well, tried, thank you anyway

[plucky pelican]

You're very welcome. I'll tell you if I have some workaround..

[lyric lotus]

Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[steep stream]

"A surreal floating library drifting through a twilight sky, each book glowing faintly like fireflies. The structure is built from polished mahogany and suspended by enormous, slow-turning gears of brass and glass. Below, clouds ripple like ocean waves reflecting the sunset’s magenta and teal hues. A lone librarian in a long coat of starlight pages stands at the balcony, releasing paper cranes that turn into constellations as they rise. The atmosphere feels calm yet otherworldly, blending steampunk and dreamlike fantasy aesthetics. Soft volumetric lighting filters through the clouds, creating a painterly cinematic composition." 📚 🪽

[fleet bluff]

A calm, elegant studio setting with soft, diffused lighting that creates a serene and focused atmosphere.
The background is simple and uncluttered, featuring neutral tones (light beige, off-white, or soft grey) to keep the attention on the subject.
A middle-aged woman is seated or standing at the center of the frame, looking confident and relaxed.
Gentle studio lights create smooth shadows and a natural skin glow, highlighting facial features softly.
Use a shallow depth of field to blur the background slightly, giving a premium cinematic look.
The camera slowly zooms in or gently pans, adding a sense of calm movement.
The overall tone is peaceful, professional, and refined — like a health documentary or expert interview setup.
Emphasize crystal-clear details, realistic textures, and balanced composition

[mortal juniper]

HELLO , somebody has a good methodology to do a perfect generated video ?

[steep stream]

"*A small Japanese café on a rainy evening, viewed through the window from outside. A cat naps beside a cup of tea while soft jazz plays on a turntable. The mood is warm, nostalgic, and peaceful.

The lighting glows with amber tones and reflections from the wet street. The style should be semi-realistic digital illustration, rich with texture and gentle light diffusion.

Captured from an eye-level perspective with bokeh raindrops on glass in the foreground.*" 🐱 🎶

[misty vale]

B-Roll Prompt 1:
Overhead shot of the figure crawling through the tunnel, light beam illuminating the walls, cinematic angle top-down.

[still hinge]

A vibrant and detailed product advertisement poster in a clean, flat vector illustration style. The scene is for a feminine hygiene product.
Main Elements:
On the left: A serene illustrated woman with brown hair tied up in a neat bun. She is wearing a simple white tank top and bright blue yoga leggings. She is in a graceful and flexible yoga pose (a variation of Mermaid Pose or One-Legged King Pigeon Pose), stretching comfortably with her eyes closed.
On the right: A bright magenta and pink rectangular package of "Period Care" sanitary pads. The package prominently features a brand logo that says "everteen" and shows a smaller version of the woman in the yoga pose.
Overall Mood:
The image should convey a sense of comfort, wellness, freedom, and gentle care. The color palette is bright, positive, and feminine, dominated by pinks, blue, white, and a touch of green. The composition should be balanced and aesthetically pleasing for a commercial advertisement.
In the foreground: A single, clean, unwrapped white sanitary pad with wings is displayed clearly next to the package, showing its shape and texture.
Background and Style:
The background is a simple, two-tone design with a soft, light pink upper half and a solid, bold magenta-pink lower half that acts as the ground.
The illustration style is minimalist and modern, with clean lines, solid colors, and no complex shading.
Stylized, minimalist leaf sprigs in shades of pink and magenta are placed behind the woman and in the corners to add a touch of nature and softness.

[livid lake]

REEL SCRIPT #1 (30 seconds - Fast-paced reveal)
DURATION: 30 seconds
FORMAT: Face-to-camera or screen recording
BEST FOR: Maximum reach & virality
VOICEOVER TEXT:
HOOK (0-3 sec):
"I cancelled $3,200 worth of subscriptions last month."

BUILD-UP (3-15 sec):
"No more Adobe. No more Grammarly Premium. No Calendly. No Hootsuite.

Instead, I'm using 5 AI tools that do the SAME work... for $87 total.

Let me show you what I replaced:"

REVEAL (15-25 sec):
"Adobe? Canva AI.
Grammarly? ChatGPT Plus.
Calendly? Motion AI.
Hootsuite? Taplio.
Video editing? Descript.

Same results. 97% cheaper."

CTA (25-30 sec):
"Want the exact prompts I use? Comment 'PROMPTS' and I'll send them to you free."

[manic ingot]

@waxen stag please head to -> #1397655624103493813 to learn how to use the bot properly

@mint rose please head to #1397655624103493813 to learn how to use the bot properly

@half citrus please head to #1397655624103493813 for a detailed guide on how to use the bot and the appropiate channels to do so

[bleak bobcat]

Incredible!

[opal cairn]

need sora code help meeeeeeee

[sour dagger]

Is there no text bots?

[glossy oxide]

As in a bot that does what our video bot does but for text?

[sour dagger]

Nothing fancy, no videos
Just generative text

[glossy oxide]

Ah, yeah can't say there is something like that.

But could happen one day!

[sour dagger]

Hmm i think just an image of toilet paper as "modern art" shall suffice
A-hem

There we go.

:p

[glossy oxide]

[lyric lotus]

Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[mental glade]

A funny short video scene:
A cute orange cat is floating high in the blue sky.
The cat suddenly starts falling down fast toward the ground.
On the ground, a small crowd of people is looking up and laughing as they watch the cat falling.
The cat’s face looks scared, with wide eyes and open mouth.
As the cat keeps falling, it panics more, then lands safely on the ground in a funny way.
Keep the style light, cartoon-like, bright colors, with clear expressions of fear on the cat and laughter on the people.

[celest yacht]

A joyful adult woman laughing brightly, playfully running in a wide, open clearing deep in the mountains of Jeju Island (Hallasan National Park). Volcanic basalt rocks, Jeju stone walls, swaying silver grass (miscanthus), cedar and camellia forest, distant oreum cones. Golden hour sunlight, warm rim light, candid moment, windswept hair and clothes, natural color grading. Ultra-detailed, photorealistic, 8k. 35mm wide-angle, f/2.8, ISO 100, 1/1000s. Composition: rule of thirds, dynamic motion, shallow depth of field, crisp focus, full body.

[heavy crest]

A cute cartoon girl in a Enchanted forest scene featuring a river crossed by a marshmallow candy bridge, trending on ArtStation, sharp focus, studio photo aesthetic, intricate details by Greg Rutkowski, highly detailed, trending on art station, sharp focus, studio

[spring perch]

please look at #1397655624103493813, same goes to anyone here expecting a video to be generated

[vernal wing]

English Prompt — same style and top-down angle as the reference image

Ultra-realistic top-down (90° overhead) cinematic shot of a real dining table.
In the center, a large white plate with a beautifully arranged dessert sits under warm light.
Inside the plate, a miniature real human chef — wearing a white uniform and tall hat — is carefully decorating the dish with chocolate drizzle, cream, and mint leaves.
Gentle steam rises from the food; light reflections move across the glossy surface.
The camera is completely static and perfectly aligned overhead, with no cuts or motion.
Ambient realism: real wooden table texture, cup of coffee nearby, spoon casting a soft shadow.
No dialogue — only the natural sound of soft sizzling and ambient kitchen tone.
At the end, steam drifts upward forming a subtle heart shape as the light fades.
Final text fades in: “Sometimes the smallest chef creates the biggest flavor

[amber hare]

Hi, everyone.
I just filled out the new LMArena survey - really appreciate how the team’s involving the community in shaping the next stage of model development and evaluation. The multi-turn editing, reasoning trace, and AI eval updates are especially exciting.

As a full-stack blockchain developer, I’ve been thinking about how LMArena’s ecosystem could benefit from on-chain transparency - using smart contracts or verifiable logs to record evaluation data, leaderboard results, or even model provenance. That kind of blockchain integration could make AI evaluation more trustworthy and auditable for both open-source and proprietary models.

If anyone here’s also working on blockchain or decentralized AI systems, I’d love to connect and discuss ideas - especially around building verifiable trust layers for model outputs or reputation systems for community feedback.

Let’s push the frontier of AI + blockchain together.

[glossy oxide]

Thank you for sharing this!! Very glad to hear you filled out the survey as well. Love that you're putting thought into this as well.

on-chain transparency - using smart contracts or verifiable logs to record evaluation data, leaderboard results, or even model provenance
Appreciate the idea!

[analog gazelle]

!

[dense ocean]

Prompt corto (para LM Arena):

Hachiro, King of the Void, sitting on a dark crystal throne, sadistic smile showing fangs, dark blue shark fin and tail, golden glowing eyes, short dark blue hair, black and silver royal coat, void mist cape, surrounded by purple-blue energy, cinematic lighting, Re:Zero anime style, ultra detailed, 4K.

[steep stream]

"A photorealistic scene of a young woman sitting in her childhood bedroom, now slightly faded and untouched for years. Posters from the 2000s line the walls, an old iPod sits on the desk, and the soft glow of a lava lamp lights the space. She’s holding a photo strip from an old photo booth, smiling wistfully. Warm, golden light filters through lace curtains, creating a tender, sentimental atmosphere. Captured with a 50mm lens in cinematic color grading. The girl says, “These are my best friends,” while holding the photos in the same way. Make everything move and slowly zoom in."

[obsidian lichen]

What are the models of the three video-generated group chats?

[steep stream]

👉 https://lmarena.ai/leaderboard/text-to-video

[steep stream]

Draw a 4-panel comic:
Comic Title: Time-Traveling Coffee
Panel 1: A person drinks a mysterious, glowing cup of coffee.
Panel 2: They suddenly vanish in a puff of smoke.
Panel 3: They reappear in medieval times, holding a goblet.
Panel 4: A medieval person says, “We call that poison,” while the modern person thinks, “Oops… wrong brew.”

[steep stream]

"A wide, cinematic shot of a ruined, post-apocalyptic Shinjuku cityscape at dusk, heavily inspired by Kazuma Kaneko's original character art. A towering demon, Metatron, is levitating above a crumbling skyscraper, unleashing a torrent of blue electrical energy. A smaller, silhouetted figure (the Protagonist) stands defiant below. Use a dark, high-contrast lighting scheme with dust motes drifting in the air, giving a palpable sense of decay and power."

[covert portal]

Hey guys, who like AI Dungeon or something like this, I have a prompt for design AI game master prompt(but on a Russian language)

[distant tundra]

A video of this man running fast, crying with tears, while shouting “Yunus Abba save me, Apa has arrived!”, emotional and dramatic scene, realistic motion, cinematic style.

[dim raft]

vertical (9:16) cinematic devotional video of Lord Venkateswara’s divine eyes slowly opening.

The camera begins with a tight focus on the closed eyes of Lord Venkateswara (Tirupati Balaji idol), showing detailed tilak, sandal paste, and intricate gold ornaments on His forehead. The lighting is soft golden with gentle rays of sunlight filtering from above, symbolizing divine morning light.

As the eyes begin to slowly open, the camera zooms in slightly and moves upward — revealing His calm, compassionate gaze that radiates divine energy.

Surrounding background: blurred Tirumala temple sanctum, soft incense smoke, and glowing oil lamps on both sides, adding a sacred atmosphere. Golden spark particles float gently in the air to represent divine energy.

Maintain a slow, sacred, cinematic motion — no sudden cuts. Focus on realism, smooth texture, warm tone, and subtle temple ambience.

Lighting: warm golden and white tones, creating a sunrise or early morning darshan feel.

Camera: portrait / 9:16 frame, slow zoom-in motion, shallow depth of field to keep focus on eyes.

[frozen yacht]

"A hyper-realistic, photorealistic YouTube thumbnail designed to captivate viewers for an Islamic suspense story about the curse of interest (riba). The central figure is a man resembling the person in the reference image—medium height, brown skin, short wavy black hair, trimmed beard—standing in a dimly lit, eerie street at dusk with a terrified expression, wide eyes, and tense posture as if haunted by an unseen force. He wears a tattered dark gray kurta, reflecting the story’s moment of despair and the haunting encounter with the shadow, with his tan shoulder bag hanging loosely. In the background, a faint fiery glow emerges from a shadowed alley, with swirling dark smoke forming a ghostly figure holding a bag of coins, symbolizing riba's punishment. The scene blends a village-like setting with cracked pavement, flickering streetlights, and an ominous sky. Use a dark palette with dramatic red-orange accents, hyper-realistic details like sweat on his face, realistic shadows, and textured clothing. Bold Bengali text at the top in red: 'সুদের অভিশাপ' (Sudar Abhishap), and smaller white text at the bottom: 'এক ভয়ানক রাতের চিঠি' (Ek Bhoyankar Rater Chithi). The composition is unique and eye-catching, with the man's emotional face in sharp focus to draw viewers in and provoke clicks."

[shell bough]

¿Model?

It's interesting for creating comics.

[steep stream]

Nano Banana 🍌

[shell bough]

Thank you. I advise putting the name of the model when sharing.

[shell bough]

New improved update of the previous protocol for the GEMINI 2.5 PRO model, now ACAC 2.0:

• Improved auto-correction with a 90-100% probability of anti-illusion and anti-ambiguity.

• It is fully comprehensive in its responses, not summaries.

• Includes self-protection to prevent long-term deviations and degradation.

• Forces the AI to use its maximum capabilities within ACAC. It is no longer a passive AI, but a proactive one, capable of automatically solving vulnerabilities without you having to ask for it in each session. For example, if you work with software code, it will be of great help to you; however, keep in mind that a manual review of your requests is essential. If necessary, perform a manual review.

• It is for global use, but works best with complex topics and tasks, not simple ones.

• It forces deep thinking, which gives better answers and prevents the AI from being vague.

• And much more.

This prompt is dedicated to people who use critical thinking, who take the time to analyze AI responses and not let themselves be manipulated by it. If you are unable to accept this reality, do not use this prompt.

Note: this will be the last update I make to this prompt; misuse is the responsibility of the individual. I have left a warning, so think carefully about what you are going to do.

[shell bough]

The IA model: glm-4.6, accepts the ACAC and does not reject it, forcing its maximum possible capacity.

[half rapids]

Ultra-realistic cinematic fantasy battle inspired by Game of Thrones. A vast medieval battlefield at dawn — snow falling, fire and smoke in the air. Armored warriors clash with swords and shields, war horses charge through mud, arrows rain from the sky. A giant dragon roars overhead, breathing blue fire on enemy lines. Epic camera movement, dynamic lighting, sparks and embers flying. Highly detailed armor, realistic blood and dust, volumetric fog, 4K cinematic lighting, 60fps, dramatic orchestral soundtrack. Duration: 10 seconds.

[steep stream]

"*A close-up studio photograph features a sculpture of a human head, possibly male, meticulously crafted entirely from dried pasta. The head is in profile view, facing left, against a soft, pale yellow or cream background. Even, soft lighting highlights the pasta's texture and creates gentle shadows that define facial features.

The sculpture is lifelike, representing an adult male with a neutral expression and closed eyes. The skin texture of face, neck, and ears is made of tiny, spherical pasta pieces—resembling beads such as acini di pepe—giving a highly textured, granular appearance.

The hair is formed by short, intertwined pasta or spaghetti strands, creating a dense, slightly messy, curly hairstyle. The entire sculpture has a uniform yellow-orange hue from dry durum wheat pasta.

Subtle details include light-colored strands forming eyebrows, contrasting slightly with the granular skin beneath. The nose, lips, chin, and ear are clearly defined despite the unconventional material.

The neck is cut off horizontally, suggesting the head is mounted on a small pedestal or is part of a larger bust. The background is smooth and slightly desaturated, emphasizing the sculpture. Soft, warm lighting avoids harsh reflections, complementing the monochromatic aesthetic of the image.*" (hunyuan-image-3.0)

[steep stream]

"The spaghetti head explodes!" (meta-ai)

[ashen crow]

A cinematic animation of a stylish sofa or living room furniture piece. The camera slowly moves around the furniture in a smooth 360° or side-to-side motion, highlighting its texture, shape, and design. Soft natural lighting fills the scene, with gentle shadows and reflections enhancing realism. Subtle depth of field keeps the focus on the furniture while the background remains softly blurred. The materials (fabric, leather, wood, or metal) react realistically to light and motion. The overall mood is elegant, modern, and minimal — emphasizing comfort, craftsmanship, and style. The animation ends with a clean centered frame of the sofa, perfect for branding or product showcase.

[urban rover]

Create a realistic image of the same woman from the provided photo, keeping her exact original facial features, expressions, and details unchanged. Place her in a live performance setting where she is holding a guitar and singing into a microphone. She should be wearing a sleeveless, shiny silver or metallic-gray dress that reflects the stage lighting softly.

The scene should resemble a live concert or small stage environment with soft, cool-toned lighting (slightly bluish). The woman has smooth, dark black hair that falls gently around her face, fair radiant skin with a subtle glow from the light, and soft, harmonious facial features. Her eyes are black or light gray, looking confidently forward as she sings. Her eyebrows are neat and defined, and her slightly open lips reveal her natural pink tone while singing.

She holds a dark brown classical guitar naturally and confidently — her left hand positioned on the strings and her right hand either strumming or holding a pick. A silver microphone stands close to her mouth. The background should remain softly blurred, suggesting a cozy indoor stage or live music café atmosphere.

Maintain a photorealistic style with faithful preservation of her identity, mood, and the overall artistic composition.

[terse storm]

“Busy Komlapur Railway Station in Dhaka, Bangladesh — lots of people walking, trains in background, realistic lighting, cinematic tone.”

[final basin]

"Scene Description:
A warm, pastel-colored village filled with nature and friendly townsfolk. The main character (based on the attached image) walks through the village, seeing people smiling and healthy. A soft breeze carries glowing particles symbolizing protection. When the character receives a vaccine from a kind village doctor, a gentle magical light surrounds them, spreading safety and well-being throughout the village.

Tone & Style:

Studio Ghibli–inspired visuals: soft watercolor palette, gentle light, expressive faces.

Uplifting, heartwarming atmosphere emphasizing community and care.

Background music: calm, emotional piano and strings.

Narration/Message (optional):
Vaccination protects more than just one person — it strengthens the whole community.”

[broken glen]

hello

hi

[sharp wharf]

Please head to #1397655624103493813 for a detailed guide on how to use the bot

[lost jacinth]

Cinematic sci-fi scene in a destroyed futuristic city at dusk. A strong athletic woman in a black tactical suit sprints through the ruined street at incredible speed, leaving behind glowing orange light trails and a swirling dust wake. Her hair flows dynamically with the wind. The ground cracks slightly beneath her feet as she accelerates. Volumetric light beams cut through the smoky air filled with glowing embers. Camera at low angle behind her, emphasizing motion blur and speed lines. The city ruins blur into the background. Teal and orange color grading, hyperrealistic lighting, ultra-detailed textures, 8K cinematic render.

[fickle cypress]

Create a realistic animation video showing a person plugging a charging cable into an iPhone. The scene should feature natural hand movements as the cable connects. At first, the phone screen is dark. When the charger is plugged in, the iPhone screen turns on with the classic Apple charging animation — showing the green circular charging symbol and battery percentage increasing. The lighting, reflections, and shadows on the phone and hands should look natural and detailed

[glossy oxide]

@fickle cypress if you're lookin to use the Video Arena bot be sure to check out #1397655624103493813

[fickle cypress]

okey thank you

[lyric lotus]

@static flicker I think you'll want to check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[steep stream]

A image-to-cosplayer prompt by the "Nano Banana Prompt Generator" tool:

"A highly detailed, photorealistic portrait of a professional cosplayer dressed as the character in the uploaded image, captured at a real-world cosplay event. The costume is crafted with authentic materials — realistic fabrics, textures, and accessories true to the character’s design. The cosplayer’s makeup, hairstyle, and expression perfectly match the character’s personality. The background shows a convention center or outdoor setting with natural depth of field and soft bokeh, emphasizing the cosplayer. Shot on a full-frame DSLR with a 50mm lens, shallow depth of field (f/1.8), and natural cinematic lighting. The final image should feel like a candid photo taken by a professional photographer at an anime or gaming expo, vibrant yet grounded in realism."

[steep stream]

"A cinematic drone shot glides over a ruined city half-submerged in the ocean. Rusted skyscrapers pierce through the mist as glowing algae illuminate the flooded streets below. A lone survivor in a tattered coat wades through the water, carrying a flickering lantern that reflects off broken windows. The camera slowly zooms in as the sun breaks through storm clouds for the first time in years." (Seedream 4.0)

🚗

[zinc beacon]

👍

[peak portal]

yousuf jolekha

make a apple tree pic

[winter plover]

Create a realistic image of the same woman from the provided photo, keeping her exact original facial features, expressions, and details unchanged. Place her in a live performance setting where she is holding a guitar and singing into a microphone. She should be wearing a sleeveless, shiny silver or metallic-gray dress that reflects the stage lighting softly.

The scene should resemble a live concert or small stage environment with soft, cool-toned lighting (slightly bluish). The woman has smooth, dark black hair that falls gently around her face, fair radiant skin with a subtle glow from the light, and soft, harmonious facial features. Her eyes are black or light gray, looking confidently forward as she sings. Her eyebrows are neat and defined, and her slightly open lips reveal her natural pink tone while singing.

She holds a dark brown classical guitar naturally and confidently — her left hand positioned on the strings and her right hand either strumming or holding a pick. A silver microphone stands close to her mouth. The background should remain softly blurred, suggesting a cozy indoor stage or live music café atmosphere.

Maintain a photorealistic style with faithful preservation of her identity, mood, and the overall artistic composition.

[coarse sage]

Create a realistic video with this prompt, pictures and company logo:
"Investing in real estate?
Consider a reputable company for professional management, diversification, and access to exclusive opportunities.
Companies offer:

• Expert property management
• Risk mitigation
• Transparency and accountability
• Long-term growth potential
  Partner with experienced professionals and unlock the benefits of real estate investing.
  Invest smart, invest with a company.
  With the lady working from the seat and presenting the prompt

Visuals should include:

• Images of properties, financial charts, and professionals
• Text overlays highlighting key points
• Company logo and branding

[lime ledge]

“On a busy roadside market, a lion-man with a muscular human body and lion face is selling hot pakoras from a street stall. Steam rises from the frying pan, sunlight glows on his golden fur, people walk behind him, realistic cinematic shot.”

[sharp wharf]

@river birch Please head to #1397655624103493813 for a detailed guide on how to use the bot

[upbeat stag]

Slow drone tracking shot moving forward through a misty, ancient forest. Giant redwood trees tower over the camera. Rays of golden morning light pierce through the dense fog. Cinematic, hyperrealistic, epic fantasy atmosphere, shot on ARRI Alexa, 8K, wide-angle lens.

[steep stream]

"*A hyper-realistic action sequence through rain-slick Tokyo streets. Sleek sports cars drift around corners, neon lights reflecting off wet pavement. The camera cuts between tire-close shots, driver POVs, and overhead drones as sparks fly from collisions. Pulsing electronic music syncs with light trails and motion blur.

Style: Cyberpunk Tokyo, ultra-detailed reflections, fast camera movement.
Mood: Adrenaline, rebellion, intensity.

Animate every car racing panel on the screen.*" (Whisk)

[harsh tiger]

A night guard working alone under a dim streetlight, checking around a quiet cell area. Suddenly, he hears a terrifying roar. A tiger appears from the shadows and starts chasing him. The guard panics and runs through the dark road toward home, breathing fast, moonlight flashing on his face, full of fear and tension — cinematic, realistic, night atmosphere, 11 seconds short film. Sad muic takbe

[minor cloud]

https://www.tiktok.com/t/ZTMDeEB2L/

[crude pike]

Create a cinematic lifestyle video featuring a man standing by a peaceful lake during golden hour, wearing a black “Fly Fishing Club – Born to Fish, Forced to Work” sweatshirt. Animate the model slightly — natural breathing motion, subtle head turn, and relaxed smile. The camera gently pans around him, highlighting the soft texture and fit of the sweatshirt.

Show close-up shots of the fabric and printed design catching the warm sunlight, emphasizing print quality and comfort. Add slow-motion footage of him casting a fishing line into the water, ripples forming around the lure.

Background: calm mountain scenery, golden sunlight reflecting on the water.

Tone: serene, outdoorsy, masculine, authentic.

Add gentle acoustic background music and natural ambient lake sounds.

End with a text overlay:
“Fly Fishing Club Sweatshirt — Born to Fish, Forced to Work.”
Available now at VibeWearTrésors.

[lyric lotus]

@unreal path Please, check out ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[sturdy rapids]

Create a 16:9 cinematic video based on this early 1930s family photograph. The setting is a sunny courtyard with ivy-covered walls and a large old tree casting gentle shadows. The group of men and women, dressed in elegant 1930s attire, share a relaxed, cheerful moment. The camera starts with a still frame matching the original photo, then slowly introduces subtle motion: soft breeze moving hair and clothing, gentle sunlight flickering through leaves, and slight natural head or hand movements as they hold their poses. Keep facial features unchanged and authentic to the original photo. The atmosphere should feel nostalgic and warm, like an old home movie or a preserved memory from the 1930s. Maintain realistic lighting, depth of field, and period-correct tones in sepia or warm black-and-white.

[stiff bloom]

Quero um prompt para uma repórter entrevistando um idoso

[lyric lotus]

@stiff bloom Make sure to read https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[steep stream]

"*The image portrays a vividly colored, animated sci-fi scene featuring two main characters set in a futuristic, neon-lit cityscape. On the left stands a human-like figure with straight, shoulder-length blue hair and large dark sunglasses. The character wears a dark suit jacket over a red shirt and has arms confidently crossed, projecting calm authority. Standing slightly behind and to the right is a tall, armored robotic being with an insect-inspired design. Its body is sleek and angular, covered in segmented plates highlighted by glowing magenta and purple lines that trace across its surface like circuits. Translucent, wing-like structures extend from its back, catching the ambient light and adding a luminous, ethereal quality.

The background features illuminated billboards and digital panels displaying stylized figures and bright advertisements in shades of pink, violet, and electric blue. The environment suggests a lively urban district filled with holographic signage and futuristic architecture. Both characters are bathed in a mix of pink and blue reflections from their surroundings, creating a visually striking contrast between organic and mechanical forms.

The composition suggests a partnership or alliance between human and machine within a cyberpunk or science-fiction setting. The human’s composed demeanor contrasts with the robot’s menacing, battle-ready appearance, implying power, control, or mutual purpose. Overall, the scene conveys strength, confidence, and futuristic intrigue against a dazzling neon metropolis backdrop.*" (meta.ai)

[modest sorrel]

Stylized infographic showing three major CCTV brands in India, futuristic glowing background, cinematic tech vibe, ultra realistic presentation of cameras without logos, professional and clean layout.

[sharp aurora]

In a serene meadow bathed in soft golden sunlight, there exists a profound peace filled with an abundance of love for the world. The grass sways gently with the breeze, and the air is filled with the sweet fragrance of blooming wildflowers. Birds sing their melodious tunes, adding to the symphony of nature. This idyllic scene represents the harmony and affection that envelops the Earth. Photo taken by Emily Davis with a Nikon D850 and a 50mm lens, using natural light and pastel-toned lighting, 8K, Ultra-HD, Super-Resolution. --v 5 --q 2

[steep stream]

"This rendering captures an audacious fusion of parametric architecture and biomorphic whimsy—a conceptual edifice that pulses with futuristic vitality. At its heart lies a colossal white prism, its facade a mosaic of jagged, polygonal facets shimmering like fractured quartz, boldly excavated to unveil labyrinthine interiors alive with sinuous, multi-tiered contours. Vast, grotto-esque apertures pierce the mass, exposing a golden-hued lattice of filigreed struts that evoke coral reefs or neural webs, where geometry dissolves into organic exuberance. A grand, sinuous staircase, edged in translucent glass, spirals through the heart of the void, bridging realms, while a secondary flight ascends to the portal below, enveloped in rippling, stone-like lattices that mimic primordial growth. The interplay of materials is electric: the shell's sleek, icy gleam yields to the innards' tactile, amber warmth, a symphony of stark lines and supple flows. Anchored at an urban nexus, it rubs shoulders with asphalt veined by lane markings, a crimson signal, verdant foliage, and fleeting figures—anchors of human scale amid the sublime. Beneath a veiled, luminous canopy of clouds, dawn's hush or twilight's glow dances across facets, the photorealism alive with chiaroscuro depth. Etched "Rui" in the corner's shadow, this vision marries machined exactitude with nature's improvisation, birthing a monument that's equal parts citadel, organism, and dream—eternally provocative." (meta.ai)

[umbral ledge]

hn

[haughty saffron]

hello

[brisk cliff]

8K stylized digital art poster of Tron: Ares, clean-shaven protagonist with long hair, cybernetic suit glowing in red and orange neon. Blend of surreal lighting and realistic rendering. Split lighting across face—half digital red grid, half natural city light. Futuristic city skyline in background, glowing text “TRON ARES,” dynamic motion blur and energy trails, IMAX poster layout.
Use this face as 100% exact
same don't make a singel pixel of this reference face from you use the exact face as provided.

[steep stream]

"These were my younger days." 👶 👦 👨‍🎓 👨‍🦱

"A wise elderly Asian man in his 80s, with weathered skin and kind eyes crinkled from years of laughter, stands confidently in a softly lit studio with warm golden hour sunlight filtering through bamboo blinds. He extends his right hand, cradling four sleek, modern smartphones fanned out like cards, each screen glowing brightly with nostalgic, high-resolution photos of himself at different life stages. The images depict him in various settings and attire: as a chubby infant (around 1 year old) swaddled in a cozy blanket; as a lanky teenager (about 15) in a school uniform, leaning against a blooming cherry blossom tree; as a vibrant young adult (early 20s) in casual jeans and a band T-shirt, at his university; and as a poised middle-aged man (around 45) in a business suit. Ultra-sharp focus on the hand and screens, with a shallow depth of field blurring the background. Cinematic lighting, in the style of a Gregory Crewdson photograph blended with hyperrealism." (Seedream 4.0)

[steep stream]

"*Create a striking 90s style movie poster titled "Echoes of the Past". The design should evoke the retro aesthetic of 90s blockbuster posters, with bold, vibrant colors, dramatic lighting, and dynamic composition. Include the following details:

• Cast names prominently displayed at the top in bold, stylized fonts
• Producer credits and production year (e.g., "Produced by XYZ Productions, 1995") in classic film credit style at the bottom
• A compelling tagline or quote such as "Every memory hides a secret" or "The past is never gone" integrated into the design
• Incorporate visual elements like vintage film grain, neon accents, and cinematic effects
• Background should feature a moody, atmospheric setting that hints at mystery and adventure
• Add space for additional details like release date, movie ratings, or logos in a retro style
• Ensure the overall look is nostalgic, energetic, and eye-catching, capturing the essence of 90s movie posters.*" 🎬 (nano banana)

[sleek pier]

First one looks like a Netflix movie

[steep stream]

A predatory expression (meta.ai)

"Generate a balanced digital illustration of an otherworldly, lizard-like visage. The skull is depicted in icy blues and verdant greens, its surface a rippling, viscous texture like living gel that undulates softly. Dominating the design are twin, elongated vertical pupils ablaze with intense amber-orange radiance, clashing vividly against the subdued dermal hues. Beneath them stretches a broad maw, parted just enough to expose gleaming ivory teeth and an abyssal void within, evoking a fierce, hunter's menace. The layout is flawlessly bisected by a central vertical line, lending the entity a ceremonial, mask-like equilibrium. Encircling this focal form is a vortex of profound navy and violet tones, evoking churning fluids or interstellar emptiness. From this milieu protrude rounded, biomorphic shapes akin to fantastical blooms or reef tendrils. At their cores, they pulse with vivid saffron and tangerine, dissolving into sheer turquoise peripheries that mirror the beast's luminous gaze. These elements drift or sprout encircling the protagonist, with grander, sharper specimens clustering at the borders and tinier clusters dotting inward, forging layers of immersion and intrusion. The ambiance fuses ethereal allure with latent peril. Stark illumination, sourced chiefly from the incandescent eyes, accentuates the tactile qualities and amplifies the volumetric swell of the wavy contours. The aesthetic merges anatomical fidelity with imaginative distortion, yielding a mesmerizing yet faintly disquieting spectacle." 🦎

[spiral loom]

A futuristic digital stage with the InterLink and ITLX Wallet logos glowing on a curved holographic wall.
Light beams form a bridge extending toward the viewer, symbolizing connection between Web2 and Web3.
Purple-blue neon reflections on smooth metallic floor.
Cinematic 9:16 frame, no people, no text or captions.
Keywords: InterLink future, ITLX Wallet, digital bridge, join the network, Web3 finance.

[lime haven]

While the sun shines uselessly, trying to compete with the polished gold of your hair;
while the beautiful lily, in the middle of the field, looks with disdain at your white forehead;

While the sun shines uselessly, trying to compete with the polished gold of your hair;
while the beautiful lily, in the middle of the field, looks with disdain at your white forehead;
Baroque style

[untold nebula]

Young Messi and young Ronaldo as childhood friends playing football together on a dusty street, warm nostalgic lighting, dreamy background, cinematic style.

[lyric lotus]

I think you'll want to read our guide on ⁠https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.

[lyric lotus]

@sudden compass I think you'll want to read https://discord.com/channels/1340554757349179412/1397655624103493813 to learn how to properly prompt the bot.