#💬│general
993158 messages · Page 621 of 994
and it caught on
Hanna and Barbara used them
that's how animation studios got so great..
Putting "Character Sheet" in your prompt might help get what you're asking for.
they learned from Disney and Hanna-Barbara
come to think of it Hanna-Barbara are prolific. I think they may have more characters than Disney
The Jetsons even are Hanna-Barbara
they were that awesome
The Flintstones
Scooby-Doo
i'd be here for a week listing their characters
at least
maybe a month
Hanna-Barbara has more original characters.
Most of Disney's stuff is copied/adapted works.
I'd agree with that
man, if they told the original story of Cinderella they would traumatize some little kids
Snow White too
and Peter Pan
Peter Pan is a demon
taking the form of a child
I have a hardcover of the complete works of The Brother's Grimm.
Those are my bedtime stories.
yeah those are the ones
the goodies
The Grimmoires
I think J.M. Barrie based Peter Pan on the mythological god of chaos Pan
with the horns that hangs out in the woods
There is another theory that Neverland is actually a place in the Feywild and Peter Pan is actually a form of Puck.
J.M. Barrie seems like he was a creepy dude
i read some of Barrie's notes he definitely made reference to Peter as a demon boy
and when you read the original story you see it
Peter comes back once each generation to take one of Wendy's descendants
so the first time he takes one of her chidren years later
way After Wendy has come back from Neverland
What I got from the tory is that Hook and the pirates used to be lost boys
they aged out of the service of Pan
Can someone give me permissions to API channel?
that's why Hook and the pirates are always trying to kill pan. he lied to them and used them
when they were lost boys
no one stays young in Neverland really
they still age there
I believe you get that once you obtain the Knight role.
SO, my guess is Michael Jackson never read the original story
The app has a lot of models, app.leonardo.ai/finetuned-models
How do I determine the modelId for use with the API?
You can generate images using the Leonardo.Ai webapp! If you've filled out the form linked in #✨│priority-early-access you can simply go to https://app.leonardo.ai and Sign Up for an account!
This is an automated message. Our apologies if this message did not answer your question.
which would be strange as he was a veracious reader
That is an excellent question to ask in the api channel when you gain the Knight role.
Does any one else have an alarm set for when their tokens reset? No? Just me... Cool 😅😅🤣
👍
LOL
Why would I need an alarm when I'm on the site watching the final seconds tick down?
i make my own emoticons
slap happy now
😎
i think this is japanese 👺
i remember the god masks they have from ancient times
Mainly cuz im at work when it does 😅🤣 so my alarm gose off i go on my lunch break and thats all i do for 30 minutes lol
Ah. It is 8 PM locally when the reset happens for me.
7pm for me
i'm an emoji machine now
If you put a link to your Leonardo profile in your Discord profile, it will be easier for others to find your profile, possibly following and liking your generations.
It looks like this without the space after the first h.
h ttps://app.leonardo.ai/profile/ your user name goes here.
king palace 13 century , river bank, mountain
@gilded vessel Yeah... We're not going to continue the discussion heading down that road in this server. Thanks.
Thanks for your cooperation.
Anyway, I'd like to say Tom Cruise makes movies for audiences
they are meant to be streamed
or watched on a home Smart TV
they were literally made for the big screen at a theater
they film is 65
that's big, people.
Cruise is a student of film. I noticed he studies even the audio production for films
every facet of filmmaking
at 61 he should know this stuff
but he learns from each film
and keeps building
learning from any mistakes and all triumphs
damn i swear i read 1x a month
He's got that thing Jordan and Kobe Bryant had
they are like Dexter in their knowledge of their craft
Different profession same drive
It's either Jackie Chan, Tom Cruise, or Stallone that is the GOAT
of action films anyway
WTH :) how do you get the Knight role? I am paying for the service and using the API for an app I am building
I don't know who you could call the greatest actor of All-Time
you'd have to go back to the beginnings to find that one
Chaplin?
Keaton?
or later?
like 1930s to 60s?
actually no i didnt misremember i have it written per their employee when i asked
depends on era
Knight- A minimum total of 15 messages, and have been in the server for over 1 day.
Thanks
You met the time requirement. You're a few messages short of the minimum.
any employees here or someone i can ask to clear this up?
How long ago did you ask that and get that response?
2w
also if i go to the pricing rn it says 1/month
most left being the free tier
It was a recent change so they could keep the free tier.
This was before the update.
@midnight cloud want to take this one?
what happened in the update?
they removed ai model training?
1 model for free tiers.
Ya 🙂 Thank you
with no refresh?
Nope.
This is for paid users.
so if i made a model under the impression it would reset and i deleted said model, now i have nothing
thats not currently updated on the plans hence why i am under the impression its 1/ month as stated
Correct.
where can i read the updates?
cus obviously this is a blindside and the app isnt updated
Don't have it handy but if you check #📢│announcements it should be there.
App isn't updated? How so?
sucks- was on leonardo only for the training of models. time to leave then
cus as of right now the app still says you get 1 a month
You could always upgrade your plan.
Training models option is not free anymore? Sorry I just heard that.
Leonardo.Ai allows users to easily create their own finetuned models. Some important things to remember about making models is that all images should have the exact same resolution and you should have around 15 images for the best results. More information about training models can be found in our FAQ at https://app.leonardo.ai/faq as well as in various YouTube tutorials. Please note free users are not able to train models at this time.
This is an automated message. Our apologies if this message did not answer your question.
wait nm they did update it- i reading old stuff
https://app.leonardo.ai/buy
On my screen it does not say "per month"
Where?
i was swapped tabs and i made that mistake-i take that back
made a mistake on that-was swapped tabs
You get one model training for the free tier.
yes i see that
All good! I'd hate to see ya leave because of that. 😦
Yeah I was replying tro Kerem.
im still going to be leaving as i have no models
sucks but it is what it is
Ahh dang, alrighty then. Do consider upgrading if you change your mind and would like to come back. 🙂
Thank you! I've tried that... didn't worked so well anyway... it was a total abomination 😅 platform models are already great for me.
LOL, at least you're honest about it! Got a favorite model?
Dreamshaper v7 is incredible.
kinda seems bs that if i deleted a model before the update it counted to the limit so my account has 0 models and wont get one. esp when done when it was under the impression of 1/month.
Thank you for your feedback and we apologize for any inconveniences it may have caused.
I love all of them... Leonardo Ai has the most amazing stuff for free. Made me forget about MJ.
I got some supercool stuff from Leo Diff especially!
Ahh that's great to hear that you're enjoying your time on Leo. 🙂 If you ever need any help or anything, feel free to reach out to us! We'd be happy to help. Absolutely love Leo Diff. One of my favorites.
Thank you so much, very kind of you 😊 🙏
I've made this with Leonardo Diffusion... I absolutely loved the result.
I was superhappy and not sad anymore because I didn't have MJ 😂
Incredible, love the mech suit on it haha
Created this for a future #1108130218260041758 🙂
@slender lance um Spice Girl/Voltron vibes
hola
But if were talking cool stuff, I made this for #1127630423988645938 😄
Daily theme: Bus Stops ? :p
I should learn how to do this and certiantly must improve my prompting ☝️ 😁
Ahhh I knew Tuxman would show up and try to get ahead. 
can you screenshot that? cus i fail to see something that says finetuned model rate has been changed. can you please give a date or anything that makes it easier to find?
If I come across it sure thing.
I see creating a model is something that is very important to you, why is it so important? Also, what can't the models available bring you that makes model training such a priority? (I'm not related to Leo... just genuinely curious)
Stone skin, hmm. Bit tricky. What're you trying to make.
cus that was literally the one thing that made this stand out to everything else.
i waited from june to july literally for a reset after testing. now its removed and gated behind paywall which seems like a total downgrade of service with seemingly 0 notice
thats why
and now i can't get a full body model to generate
just portraits
from the shoulders up
now if theres notice or something and im wrong ill gladly admit to it like above but i actually cant find anything in announcements @umbral helm
I absolutely get that you want to train a model, but for what purpose? what would you do with model training?
also the fact a deleted thing before the update means my account doesnt even have the free model seems like absolute bs.
everything about this feels ick
to make images in my stlye-like say the style i draw in
personally I could only use model training to make a proper Alita... which I've tried just as I mentioned above and results were nightmare fuel 😂 still managed to get good Alita renderings from Leo's platform models.
would be good for prototyping and if it was good i may buy for more models depending hence why i was testing - but right now im honestly more disapointed than mad and dont see myself using this service
i just want to see the announcement update before i decide if the company was scummy or if there was an announcement
Have you tried "Image Prompts" with Prompt Magic 2? Might give a similar result?
glitches out and gives 4 arms etc - do you happen to know where said announcement is?
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Scummy?

def feels scummy. like a bait and switch esp if theres no notice.
im really trying to give benefit of doubt here and looking for that announcement you mentioned.
Good morning guys
june 9 - mentions the Image to Image and Image Prompts are now also available with Alchemy are available for paid useres
june 22-To anyone who happened to miss it, please kindly note that we have an upcoming maintenance for the app, scheduled for 23/06/2023 at 3AM (UTC).
june 23-Hey everyone! Kindly note that the maintenance is now complete and the app is back online.
july 7- Hey @everyone, lot's of fun stuff to announce!
We've just released v2.0 of our Canvas.
It's got a whole heap of new features for you to play with. Note that some of these are in Beta.
This update introduces:
• Sketch2Image, Image2Image, Text2Image, and more powerful Inpainting & Outpainting
• Pixel density slider - this allows you to inpaint an area at a higher resolution which then gets resized back onto the image. It's super useful for fixing fine details like faces
• Finetuned model support for all inpainting tasks. Yes - all models on platform can now be used with the canvas.
• With the new inpainting method, you should more easily be able to add elements to a scene that aren't there.
• You also have the option to sketch what you want to add to a scene and conver that sketch to something that matches the rest of the image.
Coming Soon:
• Further inpainting improvements
• Furhter sketching improvements
• Full layer support
• Context boxless inpainting
• Other magic
Try it here - https://app.leonardo.ai/canvas
More models!
We've added Dreamshaper v7, Absolute Reality v1.6 (photorealism) and 3D Animation Style. Lykon's done a phenomenal job with DS and AR, and the 3D Animation model is super fun to play with. More coming in the next few days too!
Tweaks to Alchemy + Free Trial
We've made some changes to the Alchemy presets - Dynamic and Creative. Dynamic is becoming the new default preset - you should find it gives you more leeway and flexibility with your outputs.
Free users will be getting the ability to trial Alchemy in the coming days!
If you would like to contact us, there are several ways to do so.
-
Intercom :speech_balloon: : Access our Intercom chat directly on the Leonardo AI platform for a seamless support experience.
-
Email :e_mail: : Write to us at support@leonardo.ai with your questions, feedback, or concerns.
This is an automated message. Our apologies if this message did not answer your question.
Why are u sending it ?
@midnight cloud i dont see it in announcements- did you make a mistake in the chanel by any chance?
looking for a notice/update but i cant find it...
know when the removal of ai training models for free tier was? cus i cant find it
I think it was sometimes in late April, early May? it's been a while
so then the customer support lied to me.....
Not sure but long ago
cus that was 2 weeks ago in june
Do we have one ?
ya i have his message ss- which is why i posted from june 9
@shadow lantern Look, I see Model training is very important to you, and it just seems that Leonardo might not be the free service for you. It's great to try and explore different offers, and I hope you will find what truly answers your needs
Leonardo.Ai allows users to easily create their own finetuned models. Some important things to remember about making models is that all images should have the exact same resolution and you should have around 15 images for the best results. More information about training models can be found in our FAQ at https://app.leonardo.ai/faq as well as in various YouTube tutorials. Please note free users are not able to train models at this time.
This is an automated message. Our apologies if this message did not answer your question.
here is the msg and has 2w at the bottom
july 9 which i started summarizing the announcements is before that so any update should be noted in the frame but i cant find it
👋
hey hows it goin man
ya i know- like i said prior, i wanted to know where said announcement was
ya def seems kinda scummy
cus like i said i wanted to know if i have grounds here
there's certainly a lot of frustrations i have too but overall leonardo's one of the best AI art websites out there
i think im going now goodbye
ok! take care :D
Like I said, you are free to go explore other avenues. If you decide to come back and use Leonardo"s services, the community will welcome you with open arms 🙂
i dont appreciate being lied to espc when ive waited long espcially for a promised service that was long removed
oh ok. no problem!
i jumped into the discord at the wrong time didnt i
It's fine! People exploring and voicing their feedback 😄
Helping mom
it do be like that!
I agree with that!
Cake
😭 it's a lie too
A complex, surreal scene featuring GLaDOS from Portal, a massive and intricate AI entity looming ominously in a futuristic, sterile environment. In the foreground, a perfectly arranged, pristine chocolate cake on a clinical metallic table, a phrase "The cake is a lie" projected in ambient light on the surrounding surfaces. The scene should have an eerie yet hyper-detailed feel to it, rendered in Unreal Engine, with advanced lighting effects and global illumination. The final piece should be cinematic, full of beautiful lighting effects and elegant details, post-produced and post-processed, encoded in UHD, 8K.
Gah. I hate the default contrast and resonance. Now I have to use my fat figure to slide on two sliders instead of just one.
Why would anyone use contrast of 0.5?
What a prompt
Well, work smart. Let gpt do the work for you.
Not that I needed it these days.
Cos you can get decent image with even " " on alchemy
@velvet shuttle They keep saying that Ai can't be sentient, but now it has an imagination.
🤔
/ imagine isn't the same as an imagination. 😆
well let's say this...
The stuff I'm seeing from Ai is better than Jim Lee in his prime
An abstract, thought-provoking scene of an AI entity represented by a complex matrix of glowing, interconnected nodes and pathways, situated in a vast digital landscape. The AI, through a manifested avatar, is shown painting on a digital canvas, implying its imaginative capabilities. The environment should be hyper-detailed and rendered in Unreal Engine, filled with intricate coding sequences, algorithms, and mathematical symbols floating around, representing its intelligence. Incorporate a sense of elegance and beautiful lighting effects. The final image should be post-produced, post-processed, and color-coded in UHD, 8K resolution.
Nah. AI is stupid
You just set the template and a lot of examples.
So, you think Wild C.A.T.S. was better than this stuff from Ai?
My set up prompt itself takes three messages.
Jim is still great, but Ai sees things that humans will never think of
There are a bunch of movies and shows, depicting why trusting AI to run everything for us is a bad idea, even if it does become sentient.
I mean, it could be less verbose. My prompt bot is still lacking.
Good enough for prompt magic though.
That sad, I need to train one to give me alchemy prompt.
Parentheses will be a nice challenge. And it has to do maths now. Fun.
Do I have to unlock something in order to train a model? It just says I'm out of models. I tried one in May but it was just 3 pictures and failed. I wanted to go again and thought I had to wait a month, but it's not giving me a timeframe.
On the free plan, you get 1 model. It does not reset at the next month. If you want to train more models, you can upgrade to a paid plan.
Thank you
And yar. The cake is a lie.
Can someone tell me how to count the tokens used in a generation? I really need help with that, I'm generating with PHP, With the API
see what I mean that would make for a wild movie
maybe Sony
Pixar fell off
and DreamWorks
Thought the story is about Indiana Jones and the head of Lucas arts?
Sony is killing Pixar right now
I could use the same
with the Spiderverse franchise
Sony can kill anything as long as they make billions from cod.
Pixar is no longer innovating in the artform of animation
The amount of profit they are making. Unreal.
Nah. The past Sony is a vibrant company with many industry leading sector from music, to movie to game.
The current Sony is just a game service company that make money from whales.
guess that's why they fired all of the people that worked on that movie that bombed
They ain't even trying with their console.
Disney doesn't play
Disney is in trouble.
I have no idea how to do it 😦 do you have any guide or do you know any?
it's a cycle
As for api, I would advise against using it for image generation, unless you have to. It's outdated and the calculator that was supposed to be coming months ago ain't coming soon. So you either have to work out manually or guess.
The token use is the same as that from the Leo default ui.
f
Just that everything will round up.
So more images cost more.
Leo scheduler is cheapest.
Yes but I would like to store the used tokens in my db, I am right now with a manual calculation that I hate 😦
Basically the formula will involve image size (by total number of pixels), scheduler type use (any other scheduler will cost more than Leo) and steps (Leo default at step 10-20 depending on load. Step 30 or above will become a multiplier).
Prompt magic literally double the cost.
Hello
The highest token cost so far is a 1500x1500 image with prompt magic on step 60.
Number of images will also act as a multiplier.
Again, if you are generating for yourself, either use the apple app by the team, or use the ui.
There's not much advantage to do it otherwise, since a lot of the newer features ain't available through api call.
they are doing some weird stuff dude made an Ai OnlyFans
😄
WTF?!!!
anyway...
back to TMNT
it's going to be done in that Spiderverse style
but it's Nickelodeon
so it will be child friendly
won't be based on the ultraviolent comics
although I loved the comics too
I love anything that has to do with Martial Arts
If you wanna do ai onlyfans, then you probably wanna do it on an offline image generator.
is there a way to block the fact you blocked someone? ....asking...for a friend 😛
Not weird at all.
Well, not if the person is really observant. But basically rule is to not be talkative and respond to everyone except a certain person. Stop posting images so that the person won't be able to react.
And occasionally answer the person so that it seem like he/she/they are not blocked.
Show them u don't blocked them
Hey Electron!
This is what I get when I prompt for Teenage Mutant Ninja Turtles
😄
it's a believable image though
looks real
just not a turtle
Its the seeing how many blocked messages I have missed because I blocked because they posted too many in the first place instead of saying everything at once. LOL... AFTER being asked by mods to NOT do that...LOL
I guess that would be called a blocked message notification- but I cannot find anywhere in setting to stop that particular notification. Hahahaha
Just ignore it. They will get it out of the system. And soon its all peace.
Like a.msg in Chay saying blocked msg ?
except her right eye is deformed
other than that believable
it's just the pupil of the eye
but still
i noticed
@boreal turtle share a ss of what u trying to say
Try turning on prompt magic. Or put teenage mutant ninja turtle at the start of the prompt. Order matters.
Turn on prompt magic but set the init strength to the minimum. No need to add reference image.
anyone got experience fixing this archer's bow? Its so hard to get it right
its like sprouting everywhere from its body
Do you have it prompted to place the bow over one shoulder?
Good luck with generating bow. @hidden galleon has lots of experience trying to make bow work.
I can't even make a figure take a bow most of the time.
my prompt supposed to get him hold the bow and draw the bowstring but all rubbish
yea ... so damn hard
You obviously don't get the joke.
But yar, bow is not an easy object to generate on the models.
I was able to get mine 'where' I wanted just need to get it in one peice! 😛
oops... sry..to focus on getting it right.
what did u do there? T.T
I prompted: Holds bow in front of her arrow drawn and ready.
It took a few tries to get one close and then i Image to imaged her. NO prompt magic (it cuts the heads off for me 😛 ) and I did not use alchemy either.
some chick from an alternate Earth with an indigenous man
Spoil it
yea alchemy and PM screw up my prompt too.
she exists somewhere
I'd bet money there is a blonde that looks like her on this planet though
Please put a spoiler on that kind of stuff.
Some people get excellent results using one or the other or both...I am just NOT that person! hahaha
🫡
Alchemy + pm ain't hard to figure out. Expanded domain though. Haven't gotten anything consistent with it.
Can you add a real photo to canvas and then edit it with a prompt for the person to be holding a broom or something? a cup coffee
ya know
?
they arent bad...but usually end up taking my 'idea' off into never land lol
like say I take a phot of me and make myself hold a baseball bat
something like that?
or change the outfit
Sounds more like a trained model than prompt magic.
Leonardo.Ai allows users to easily create their own finetuned models. Some important things to remember about making models is that all images should have the exact same resolution and you should have around 15 images for the best results. More information about training models can be found in our FAQ at https://app.leonardo.ai/faq as well as in various YouTube tutorials. Please note free users are not able to train models at this time.
This is an automated message. Our apologies if this message did not answer your question.
Heh. I have yet to train a model.
yeah i can't get my comic book cover to generate a character page
or add characters to the cover
or anything
won't edit the image in canvas with prompts
so idk
Alchemy + pm. No image given.
@midnight cloud sir bug or what ?
Same prompt without pm. Just alchemy.
The prompt:
An engaging and epic scene depicting a fiery orb, or fireball, sitting alongside a piece of tantalizingly detailed cheese. The fireball illuminates the cheese, casting intriguing shadows and emphasizing the cheese's texture. The scene should be rendered with Unreal Engine, giving it hyper-detailed attributes. The flames dance with beautiful lighting and global illumination. The elegance of the composition should be emphasized in post-production and post-processing, resulting in a color-coded UHD, 8K resolution image.
Basically pm is good at interpreting to the word.
Imagine/ one woman wear bra and reyal skin fully body
Leonardo can only be used through the webapp. You can get access to it by filling out the form linked in #✨│priority-early-access!
This is an automated message. Our apologies if this message did not answer your question.
Reyel?
You have touched a challenging topic....it seems Bow is not easy to generate, so is Axe..... Some use canvas and sketch to add it in...some use image to image to prompt it in
Imagine/ lod Krishna image reyal skin. He's swear jewelry
Yup, but for person holding axe etc....that model not sufficient I think
I mean, controlnet is still the most reliable way to get the object interaction done.
I must say that the recently added models are way better with holding of objects too.
Provided that the object can be summoned.
This is with Prompt Magic yea?
I posted both images with or without pm above.
Oops. Got it.
This looks like barbeque apple with a side of cheese 😆
No preset for obvious reasons.
wat does controlnet do? i understand it on works with img2img but how does it infuence the image
#1122654030808748153 check this
ok so this is where the guide or manual hides
Leonardo.Ai is a Stable Diffusion based AI. This means prompting techniques applicable to Stable Diffusion should be used in Leonardo.Ai. Do note that Leonardo.Ai does not yet support weighting and that unlike Midjourney there aren't any prompt parameters.
If you would like to know more about prompting, there are various guides linked in #📌│faq as well as tutorials you can find on YouTube. Happy prompting!
This is an automated message. Our apologies if this message did not answer your question.
@tough summit #1093657915740672041 this thread is also fun....to join in
i know you are hungry coz its lunch time
Is it too late for me to answer that? 😄
Tuxman? it is NEVER too late for when you answer stuffs! 😛
You can answer...because I'm always lurking 👀
Long answer (sorry)
Basic image generation normally starts from a random colorful blur that will get changed and modified at every step to get closer and closer to what was asked.
Basic Image to Image will create a blur based on the provided image. So the generation will keep the Colors and tints roughly intact
(Ill continue with ControlNet in just a sec 😜)

I like how you avoid the word noise.
Switching from Dalle to Leonardo.ai, so far super happy with it
I mean, it's a diffusion model. It's all noise.
ControlNet will take the provided image but change the “initial state” to guide the generation.
-
Pose to Image: this one is the most similar to normal Image to Image but it will ignore Colors. It only takes a blurred “grayscale” picture to start the generation
-
Depth to Image: it will take the base image, and create a black and white version to bring out the “textures” and well… depth… to reconstruct objects around these shapes.
-
Edge to Image: it will trace the outlines of the base image (think a bit like a colouring book) and will try to fill the colors and textures within these “edges”.
So they are in the order they will be from looser to more adhering to the base image (but of course you have a init strength you can play around with to give more creative freedom to the program)
Pin this....
-it’s noise all the way down?
-Always have been
🧑🚀🔫🧑🚀
I try doing edge to image based on a photo of my daughter, but it always comes out all wonky
The outlines are not defined enough
Yes this is what I suspected....
The photo itself not suited
Take a picture in front of a contrasting wall Color)
Exceeding the character count I believe.
Finetuned model will work better for face swapping.
Ohhh
Anyway, goodnight all! 😴
Controlnet still relies on the pseudo randomness in generative ai
If only I can get her to sit still 🤣
Good night Frap
I mean there's a reason why all the ladies for preset look pretty much like the same person with different wigs
Anyway thanks for this. I have taken notes, just to let you know it doesn't get lost in a sea of messages on #General
/imagine prompt: Make an anime background image: A Chair of an king in the middle of the room, Night theme, Moonlight coming from window coverings half of the chair, Lanterns, Walls have bricks structure, provide only outlines with black marker::5 anime::5 --v 4
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Well, mostly cos there are a lot of ladies images generated.
I always suspect it’s like the “mean” face from all assembled training data for that model 😜
But still. They look pretty much like the same person.
Hi
Okami
“Mathematical mean” .. I mean… don’t be mean
I mean the default image to generate is still a white lady right?
Yar. I know what you mean.
AR model has a more asian twist in flavour for the original character.
Less anime like ds6 and 7 though.
Denise looks like “undefined general Mediterranean region” 😅
I had the chance of getting these with my daughters 😜
ok now i got it
R. Kelly if he were born female
I didn't think it was possible, but it is
😄
it worked
i have diff poses
Yeah you showed me before. Beautiful....I'm trying to covert my daughter to a Supergirl....but I need her to stand still and not running around pretending to be one 🤣
Oh. Ds6 and ar works well with rule 63
Someone has been doing IG models this week
Aaaand… Depth to Image 😜
hmmm.. i don't see Kels in the face there, though
that looks more like that chick from saved by the bell
the bipolar one
The chest ain't right
Oh absolutely! Like I said, Depth takes in the “textures” and the bumps and reconstruct the image. It takes creative liberties, so to speak 😄
Is there a channel for
apps created with the api
Hello! Leonardo is a webapp, not a Discord bot. That means commands such as /imagine, /image, or similar will not work in this server. You can get access to the webapp by filling out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
There is #💞│self-promotion if it’s to showcase it to a wider audience
Or #💻│api for questions and discussions about it
cool thanks - looking for stuff others created
There is a couple of app floating around on #💻│api for image generation. You can search around there.
But basically I don't think generation is the way to go with the current api state.
It's better to look at scripts floating around to better manage your image gallery instead. Ain't apps though, more like codes.
But it is a lifechanger compared to the current navigation of the super slow db
Should have used AR instead of DS7 😅
This chest is better.
But the clothing don't make sense
Is it tight or loose? Make up your mind.
I was kind of going for an art piece kind of look
instead of photo
Hey! It’s called fashion!! 🤣
AR doesn't mean photo.
like painting or color pencil-ish
Stop thinking that model decide what you type of image you get. Doesn't work that way.
AR means real AF
Like that would be actual Leonardo Da Vinci
Da Vinci was so gifted his paintings looked like photos
long before the camera was invented
If I find it will post
yep and how that fabric works in the real world. 😛
This is an AR image
@velvet shuttle that's some hard shiz. Matrix reimagining?
yeah I like it
When is deliberate 2 coming??
If the waterfall is Matrix code then hell yeah
it looks like code from what I can see in the detail
neat
When you ask for mlp and the model says no.
Yar. Definitely need to retrain my gpt bot for alchemy
Oh! It misunderstood for Equestria Girls
But I asked for pinkie pie and fluttershy
Nice
I just realized you got the Same dress Colors 😂
i'd love to try leonardo.ai
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Neon wolf with moon
Poor wolf
How are y'all able to create this?
By using words.
I've been using words but the outcome are always bad 😞
Can you just explain every selection
What are you trying to create?
W o r d s is a English term
how do i correctly define an anime style of a specific (typing an artist doesn't seem to make a huge difference)
Okay explain
Ahh use gpt
Also, sometimes with anime it helps to use the name of the studio..and not just the artist...as each studio has it's own style with anime artists used to re-create it... like using ,"In the style of Studio Ghibli" helps mine....
yvw! Let me know if that helped! there are other studios too, Toei ANimation, MAPPA, Kyoto Animation, Studio Pierrot to name a few
Let me.know if didn't work
will do!
also if there is a manga or anime that you really like? You can check to see which studio created it and try using that tagline.
i did that, it kinda did the same style the artist did but i'll just try to tinker around with it
i'll ask for help again if i don't generate what i wanna generate for 30 minutes o so
Leonardo can only be used through the webapp. You can get access to it by filling out the form linked in #✨│priority-early-access!
This is an automated message. Our apologies if this message did not answer your question.
/
i can't post images for a comparison
We have a channel where you can post and view images generated with Leonardo: #🖼│image-share. To post in it, you will need the Knight rank which you can get by being in the server for more than a day and having sent 15 or more messages. More information about ranks can be found here: #📢│announcements message
This is an automated message. Our apologies if this message did not answer your question.
you need at least 15 messages sent to gain the knight rank...and then you'll be able to post images
Ahh, thanks
It just started, i'm currently at school right now
Shall we ask mod for role for her
YUS Dude!
it's gonna be for a comparison for what i'm aiming for and what i'm getting constantly
gandalfG seems to be a good one.. 🙂
Nice!
@lime crag can @lucid hare have knight role
all ready have that link but thank you
BRB, i'm gonna grab a mouse real quick
YOU'RE a Knight!
what are you trying to do billy?
TYSM! okay here we go
left is the result, right is the style i'm trying to achieve (maybe i need to be more specific)
maybe i'm using the incorrect model?
left result is nice though...But I can relate to it not being what you're going for!
I've had good results with DS6 and Deliberate
i'm using ds7
I tried several times to draw an archer, it's amazingly difficult. This is about as good as i could get from a prompt alone: https://cdn.leonardo.ai/users/45b89e85-efbe-4a7b-9e2e-8a9195a577d5/generations/fe4e0c76-063e-4171-b23c-21049ea3839b/RPG_40_A_determined_archer_stands_atop_a_hill_bow_drawn_and_re_0.jpg
I have had excellent results with DS 3.2 too
Some times...new isnt always the best. 🙂 And each model has it's very own strengths..I havent figured out wht those are yet...but I'm working on it! lol
haha
Both are cute
yeah ds3.2 is more what i'm aiming for
One month i went through each model and created an image using hte same prompts etc...and created 2 images per...and put them in a folder os i could go back and look at them. For what each one seemed to offer
absolutely agree
Hehe
did you use an actual bow type in the prompt?.... not questioning your abilities just curious. lol
when i did mine I specified long bow and it seemed to help tremendously...not perfectly? but it helped. lol
it feels a little bit better now
now i'm gonna try deliberate
what if i use image prompts?
Why is Stable Diffusion (either version) no longer in my list of models and how do I find it???
Image prompts help me a LOT when im going for a specific look
i'll try them
did you try clicking the Select Custom Model option? Mine are ALWAYS changed when I log in and the ones I use most are never there LOL..so i have to reselect them.
But when I do that, where do I find it? I can't find it anywhere under "Platform models." Is it under "Community models"? "Favorite models"? "Your models"?
I can't find it anywhere!
The only Stable Diffusion ones i know of are Leonardo Select and Leonardo Creative and Leonardo Signature
L Signature is based on SD v1.5, L Creative is SD 2.1 and Leonardo Select is SD2.1
Any idea why a change to one of my works won't save in canvas?
an edit anyway
come to think of it I don't see a save option or hot key for saving progress
The buttom button on the left
Well, it's still in the Canvas, and it looks like this.
Sorry, I when you said save, I thought you meant download.
To save your change from a prompt you need to click "accept"
Wow. 👍
eh... what checkmark in canvas?
there is only Snap
and that is not a save feature
I thought saving in canvas mode was not an option?
seems like that would be correct
because it is not saved in my personal feed
the image is unchanged there
but the image is edited in canvas
hiya'
already did that
not saving
Of course. Because you can add in a number of images onto canvas to merge for example. Makes it difficult on the backend to know what to create and post in your feed in that case.
All images you generate are viewable by all Leonardo.Ai users through the Community Feed. Users who have subscribed to a paid plan can make images private by toggling the "Public Images" toggle in the sidebar. The toggle only applies to images generated after toggling it.
This is an automated message. Our apologies if this message did not answer your question.
How do I shut the bot up?
my changes are not saving to this image
that simple
no point in me altering the images if I can't save the change
Saving to your gallery is not possible.
But downloading to your device and saving your changes from prompts is.
and when I try to edit an existing image and upload it from my Lap i can't alter it
so weird to me
dont use any triggers. 😛 Otherwise we are stuck with it. lol Since this is general and not the chat channel
It may look simple. But your feed is currently only fed with tables from the image generation info. The images in the little box in canvas is actually saved in your profile. Just be thankful that they don't go into your personal feed.
Generating new entries otherwise ain't coded at the moment. Not a wise thing as it can be easily abused.
So live with it like a canvas. And then keep it in your own collection.
abused how?
that's just it i want the image change to be saved across all of my account
not just in canvas
If you really need to have it in your profile, use the final image as img2img with near full strength.
soon as I leave canvas I'm screwed the image will revert
You need to download the image to have it saved.
hola
Abuse how? I dunno. How does face swapping sound like? Or making it erotic without the wording moderation?
can I put in image in here, can't in #🖼│image-share
We have a channel where you can post and view images generated with Leonardo: #🖼│image-share. To post in it, you will need the Knight rank which you can get by being in the server for more than a day and having sent 15 or more messages. More information about ranks can be found here: #📢│announcements message
This is an automated message. Our apologies if this message did not answer your question.
Canvas can not be saved into the app itself you must download the image.
I just need a few more messages then
You unlock image-share after sending 15 messages total
I am figuring how to make the best looking generated images I can get Leo to make
that will happen regardless. LOL and word modification is l allredy absurd and doesnt even work. As an old lady...nothing surprises me any more Sorry if I triggered something with my simple question
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
how can create a imagine with ia ?
Hello! Leonardo is a webapp, not a Discord bot. That means commands such as /imagine, /image, or similar will not work in this server. You can get access to the webapp by filling out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Those fine lines really annoying. I get alot of those too. 💢
Remember, not everyone grow up with file structure and folders as second nature. I learned it the hard way when I deal with gen z.
I tried Canvas and was confused how to use it lmao
You can generate images using the Leonardo.Ai webapp! If you've filled out the form linked in #✨│priority-early-access you can simply go to https://app.leonardo.ai and Sign Up for an account!
This is an automated message. Our apologies if this message did not answer your question.
does anyone else see the bot or is it ust me
Nope we all see it 😛
We all see the bot.
ah
Sadly everyone sees it @lunar pasture
sadly
Yes
hey hey McWurth!
Hey VinRey 😊
I mean, abuse will happen. No point making it easier.
The point is more that the current implementation of the db is more concerned with the generation info and result, not placement and overall canvas image.
It sometimes slips my minds that the grandparents of today helped to build the digital universe in the 1960s, 1970s and 1980s some in the 1990s with Sony Apple and Nintendo
not to mention Sega
It's not hard to include it in. But the pros doesn't outweigh the current state.
THAT is Truth!
Something about make a female redhead in armor is something I love having AI generate for me
Hello! Leonardo is a webapp, not a Discord bot. That means commands such as /imagine, /image, or similar will not work in this server. You can get access to the webapp by filling out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
like battle armor
@lunar pasture are you Norwegian?
Rapunzel realistic image
Ah ok
oh lawd a neighbor!
hiya
Name is actually more Icelandic
the S is Czech
I'm happier with the canvas not on personal feed anyway. This way the prompts you see on the feed are more likely to be reproduced.
or Eastern EUropean
Again, not saying you can't punk around with the prompt.
I am sure I have sent 15 messages so faR
I love it out of the way as I never use it 😛
imma ust make my pfp the image Leo generated
Welcome to the Leonardo.Ai community!
Leonardo.Ai is a web app (not a Discord bot) which you can use to generate images.
To get whitelisted, please fill out this form: https://bit.ly/leonardoai_earlyaccess!
This is an automated message. Our apologies if this message did not answer your question.
I Love mcWurths! and mine was created by my brother in leo too
What is McWurths
Thanks. It's from a series where I used same prompt on all alchemy settings with alternate reality.
the other person talking to you?
ugh can I just post like 3 images in here
You can post images in #🖼│image-share once you have the Knight rank. You can view it with the Novice rank (being in the server for more than an hour and having sent 5 or more messages).
This is an automated message. Our apologies if this message did not answer your question.
I toy with alchemy but I havent created anything I actually like with it yet. 😛
And Imma token prude! hahaha
I'm not even using alchemy
kept expiring
I love this one
Try lower down the resonance way and up the contrast.
I am not. Some generations cost me a LOT
Nice.
@lunar pasture By any chance is that character from Canada?
Nah
Lol no way... alchemy is a game changer
probably could find a way to remove it via canvas
think of it as a weird umbrella corp symbol I guess
I keep hearing that and 'seeing' it...but when i use it?... Im like..."Seriously!" lolol
Maybe need to play around with the preset
Ai is making up planetary alliances and countries
like Star Trek
probably. I'm trying to ration my tokens though :P... so I will practice my prompting and learning the different models frist...THEN I will play with all the little doo-dads. 😛
Left is without Alchemy, right is with Alchemy
Walter white with the peter griffian chin
both are nice ...but I have to say I like the coloring/ vibrancy of the left
Yeah because I used the Photography present on alchemy...didn't really try the others lol
looks like i never need alchemy
Forget preset. Alchemy is godly because of the weights.
so all is well
I mean, preset are good.
I only started testing it since it's launched on free accounts
But you wanna have fun, you gonna play with weights.
Still learning
I like how in this one it makes eveyone a redhead
still don't know why it did that with the jacket though
#🔰│getting-started photo of boy
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
um... @lunar pasture is there a Black Widow fanclub convention in your scene?
I like trying to make realistic images...
#🔰│getting-started boy playing tennis
Welcome to the Leonardo.Ai community!
Leonardo.Ai is a web app (not a Discord bot) which you can use to generate images.
To get whitelisted, please fill out this form: https://bit.ly/leonardoai_earlyaccess!
This is an automated message. Our apologies if this message did not answer your question.
looks like a woman I saw in a movie
like last week
same and it usually does look REALLY good...then i get things like the monkey fingers... or fingers showing up wher fingers have NO rights to be! lol
Lol
man the 80s fantasy aesthetic hits different
https://youtu.be/261DaIbkRac
when they get that weird fog thumb
Try absolute reality @boreal turtle it is pretty decent with hands
Absolute reality is great
nice riddle in your profile
#funny movies
hmm i will give it shot!
whats a dataset
@celest vessel can you stop please and read #✨│priority-early-access
is that a WWE chick?
A dataset is a set of images you can use to train a finetuned model!
I'm like damn she looks real
No, it is a character I created myself. @gilded vessel
my guess people presume you're a woman
but I always investigate first
Gave a name and kept changing the name till I was happy
Yup. I'd been using nothing but AR for last few days. Best anime model!
I still waiting for deliberate v2 for stuff I really wanna do.
@velvet shuttle it's really cool anime+ar1.6+alchemy
so I grab images and the dataset uses those
Get preset out of here
I like this
I did that at first as well but tweeking the settings is really fun to do @gilded vessel
that's going to make a great video game character
or hell make a movie
get Sony or Disney's attention
Maybe save...
Disney's
2
im skeptical for the kraven movie ngl
it could be good but im expecting it to be mediocre
oh and Venom came to Spider-Man's universe the Tom Holland Spider-Man
IT won't surpass Mobius
it's about to go down..
that's for sure
LEMME TELL YA SOMETHIN'!
My boy at Comics Explained is hilarious with the catchphrase
😄
It's gonna be too good no one knows who Peter Parker is anymore because of the Dr. Strange spell
so this is going to get real interesting real quick
Then the X-Men are coming back eventually
it's on
well Morbius isn't in the MCU
he is now
how
Venom got sent back to his universe but left a sploodge of himself
still, Tom Hardy venom got skadooshed back to Sony
I thought you got this out of your system last night? lolol
im more excited for beyond the spiderverse if im being honest
a new venom most likely
Spider-Man in is the MCEU
but Sony has the rights to SPider-Man
and all of his characters
that's why you see only SOny making Spider-Man movies
fancy? can you Hit ME over the head with your hammer please?
LMAO
in the comics there is a venom that takes the host of a hispanic guy, and venom left his sploodge at a bar that has a hispanic bartender
but Sony works with Disney and Marvel obviously
snicker~snorts
that's how Dr. Strange got in the Spider-Man movie
didnt disney finally get the rights to the hulk?
You own the full rights to any image you generate, which means you are free to use them commercially. However, copyright on existing logos and characters still applies, same as if you had painted the image yourself.
This is an automated message. Our apologies if this message did not answer your question.
shush bot
ha
not to make a solo movie
dont get him going...or we will be stuck in a 3 word pings for DAYS
cause its finally on disney plus i think
contracts man
The MCU is broken.
not the 2004 movie though
😦
SONY got a great deal when they bought the rights to Spider-Man
SPider-Man is the most beloved MArvel character of all-time
he's like the Marvel GOAT
the change ┬─┬ノ( º _ ºノ)
The old Spiderman yes.
i think he was the character that launched Marvel
it was Stan Lee's first Marvel character if I am not mistaken
in terms of popularity hes kind of like marvel's batman in way
:wave: Welcome to the Leonardo.Ai community!
To use our webapp to generate images, please fill out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Yes, Peter Parker in every universe. except Gwen's universe where he died from making himself a monster (The Lizard)
You know what is funny, most people nowadays don't even know DC stands for detective comics
he co-created captain america in the 40s i think, back when marvel was still called timely comics
what happened to the smooth, crisp, and alternative upscale?why is not working?
im pumped to see what james gunn does with the rebooted dcu
i have faith
I honestly don't bother anymore with the new movies. It was fun for a while but now it's just meh
they really put the detective work into how the DCEU is so awful except for like 3 movies
sike
i mean
man of steel, wonder woman, and justice league snyder cut are leagues above most of the other movies
maybe shazam
oh and the suicide squad remake
The DC movies are just splintered if someone can unite the main Universe in movies
Suicide squad reboot is good
It's all there on the images?...it works...sometimes takes a few seconds to hit' though
then we're good to go
Imma take a shower
the stories are all standalone
well james gunn has like a 10 year reboot plan
they go nowhere
so we'll have to see
and he's going into more obscure stuff to really make it feel unique
Marvel connects everything
that is the beauty of it
the beauty that Dc mastered in their animated movies
The Leonardo.Ai Discord server has a rank system based on activity in the server. Higher ranks unlock certain channels within the Discord. You need the Knight role to post in image-sharing channels!
Ranks:
:orange_circle: Novice - 5+ messages and joined over 1 hour ago.
:green_circle: Knight - 15+ messages and joined over 1 day ago.
:blue_circle: Master - 100+ messages and joined over 10 days ago.
:yellow_circle: Noble - 1000+ messages and joined over 25 days ago.
If you have questions about a specific rank, feel free to ask and I will answer to the best of my ability!
This is an automated message. Our apologies if this message did not answer your question.
DC animation can't even be touched
well, not as much post endgame imo
true, mask of the phantasm is 11/10
I was like......
as much as keaton's batman is amazing
Did you find it?
no fracking way he's doing Gods and Monsters right off the bat???
With a Wonder Woman that is a New GOD and lives on Genesis???
The Vampire Batman
and Zod is Superman
???
though they are going straight to damien wayne being robin
Lets try and put our words into one spot. lol you guys are pinging ALL the people left and right
there is an animated version
that shiz was niiiicee
I was like what Batman sucks blood?
perfect for him
seeing as he fought Dracula is a book
yes Batman defeated Dracula
that's crazy skills
YO i forgot they have a swamp thing movie in PHASE ONE
he can beat anyone with preparation lol
apart from like, starro
they fracked up when they let the streaming series get cancelled
Swamp thing the series was great checkout the 1st and only season
that was going somewhere
I hate NetFlix now too they cancel all of their great series
Lovecraft Country
awesome
Jupiter's Legacy
they got back all of that tax money from the gov cancelling shows
DUDE
but didn't have enough money to produce another season of Jupiter's Legacy?
like wha?
Like you DO realize every time you hit enter before finishing a thought that you PING everyone listening for people thatneed help?
okiez
Im only saying becasue you were reminded yesterday..lol... Not mad
BTW.. there is a new theory that time is broken, but time is an illusion. It doesn't exist. So, not sure how it's broken.
why is Chat GPT making such great prompts for me that are like 100x better then me and Leo actually makes them look 10000000x better then what it gives my prompts
Leonardo.Ai is a Stable Diffusion based AI. This means prompting techniques applicable to Stable Diffusion should be used in Leonardo.Ai. Do note that Leonardo.Ai does not yet support weighting and that unlike Midjourney there aren't any prompt parameters.
If you would like to know more about prompting, there are various guides linked in #📌│faq as well as tutorials you can find on YouTube. Happy prompting!
This is an automated message. Our apologies if this message did not answer your question.
AI ineptism
favoritism?
yussssss!
We have achieved ascension
that is it
This character needs her own movie
video game
and possibly comic
you gotta make it happen
stop all ready
name her Aloy
Movie Movie Movie
oh wait
i noticed they don't make video games based on movies anymore like the 1980s 90s 2000s anymore
yeah
it's the other way around
imagine a open world game where you play as The Crow
Imagine/ cat walk in the moon
Hello! Leonardo is a webapp, not a Discord bot. That means commands such as /imagine, /image, or similar will not work in this server. You can get access to the webapp by filling out this form: https://bit.ly/leonardoai_earlyaccess
This is an automated message. Our apologies if this message did not answer your question.
Resident Evil
i would kill for an actually accurate resident evil movie that leans into the retro horror vibe of the original game
Uncharted
I knew this would happen. I had been waiing for the day when tech would make games look like movies
now everything is cinematic in games
what is the word limit for Leo
1000 characters
This is a great time to be alive
More like 1000
ah 1000
Yar. I see what you did there
1000 characters and anythign above 75 words and Leo starts reading blah blah blah blah...LOL
Unless it's prompt magic.
The weightage is different on pm.
It doesnt decline as fast
Oh. One key thing to note. The images are trained based largely on tag information. So any info beyond 2020 will largely be nonsense to the model. It will guess, but your mileage will differ.
So don't expect to get something like oshi no ko from the models.
Unless it's a custom model.
Hello everyone, I am Jashan and I am developing an application using leonardo AI API. Currently I have only three options for the models in the documentation. But I want to use the dreamshaper model. Any suggestions on how I can do that?
@shadow lantern could you please kindly send me an direct message. Thank you!
It should be possible to do an API call for the dreamshaper model
@karmic spoke would you be able to assist when you're around? Or @sharp zenith for that matter. Thank you!
Ahh ya sure
I'm here!
Go to fine time model, under dreamshaper, click find out more. The info should be there.
Finetuned model
I can help
Ig u need to use dreams shaper right? @misty vortex
Any resources for the same?
Leonardo Creative: 6bef9f1b-29cb-40c7-b9df-32b51c1f67d3
Leonardo Select: cd2b2a15-9760-4174-a5ff-4d2925057376
Leonardo Signature: 291be633-cb24-434f-898f-e662799936ad
Currently these are the only options in the documentation
Ah
for these
I got this 😅
Here you go
This way you can see the ID of any model
Yes, I need to make an API call to it
Yup click on the model u Wana use and then click on view more and u can get that
sweet thank you
Thank you so much
No problem, here to help 
Didn't update since April, cos there's no point in generating images with api as is.
Let me know if you have any more API questions, I have quite a lot of experience with it :)
But with API u can do cool stuff like
!enablemodule generate
!generate a beautiful cat --promptmagic
LoRAs will thankfully help somewhat
!disablemodule generate
How wonderfully slow
That is Discord
And u chose the imagination war here