META MJ - 🔥 MidJourney Image Prompt Creator 🔥Super Easy - Super Stable - Awesome Results!

[drifting coyote]

This primer prompt provides: a highly interactive, user-friendly and dynamic process for creating a custom image prompt for the user within the ChatGPT system.

The system: relies on user inputs, preferences, and questions to refine the output and tailor it to the user's desires.

The primary goal: is to create an image prompt that closely matches the user's vision, using vivid and creative description vectors. The user can provide input, ask questions, and modify aspects of the image prompt, while the AI actively adjusts the image vectors based on the user's preferences.

** Features:** The prompt offers an interactive and flexible experience, allowing users to provide input, ask questions, and modify aspects of the image prompt. Customization is key, with users defining the subject and setting, adjusting tokens and parameters, and using priority indicators.

** Benefits:** 🎁 The process results in personalized output tailored to user preferences and focused on the desired subject and setting. Enhanced creativity is achieved through vivid description vectors and AI suggestions. Users have control over various aspects of the image prompt, including aspect ratio, quality, and excluding specific elements.

**Super Powers:** 🚀 The AI actively adjusts image vectors based on user input and preferences to perfect the image prompt output. Robust and interactive questioning enables the AI to gather relevant information. Tokenized and default values for parameters, such as aspect ratio, image quality, and image weight, are used to ensure proper formatting and placement in the image prompt.

 **Requirements:** ChatGPT-4 and some creative juice

DEMO Video Here: 👇

**Awesome Image Results Here: **👇

**The META MJ MidJourney Image Prompt Creator:**👇

[drifting coyote]

Here's a few of the DEMO images 👇👇

[reef jolt]

woah this is awesome

[drifting coyote]

Thank you for the feedback. Please come back and share your thoughts so I can make it better for you. Also, share some images here that you made 👌

[tepid shale]

That's extremely useful! I also liked the documentation you made above. It is so helpful to have explanation for the prompts.

[coarse crescent]

This looks very good. 👍

[drifting coyote]

Thank you for the feedback @tepid shale ! I am working to make my prompts professional, informative and user friendly. My goal is to make AI tolls that make AI more accessible for any user.

[drifting coyote]

Thank you @coarse crescent . Please come back and share your images, questions or feedback. This helps me make it better for everyone.

[stray socket]

Super cool, Thank you @drifting coyote I am ready to use it for bringing more art to life, Would be best to have a version for the OpenAI Chat Playground where User can provide all the parameters at once?

[drifting coyote]

I'm finishing a version that works that way 😎

[lofty drum]

Super cool! Pity that I only have ChatGPT3.5

[drifting coyote]

I am still working on a version for 3.5. The issue is how much GPT needs to know about MJ. 3.5 doesn't interpret the same wasy as 4 annnnnd it has a difficult time in a conversation that starts with a bunch of tokens.

[willow zenith]

Great work! 👌👌👌

[drifting coyote]

Thank you. Please let me know how I can make it better.

[valid trail]

Amazing!

[drifting coyote]

I'm happy you found it useful, please come back and give feedback so I can make it better. Also feel free to share your favorite images you create with it here.

[mystic wind]

This is good but if you can make it work for 3.5 somehow, that would be great

[drifting coyote]

Definitely working on that. The challenge with MJ prompting is due to how much GPT doesn't know about MJ. You have to give a ton of factual data for it to understand all the different parameters and output context.

[bright idol]

Awesome work!

[vivid silo]

Thanks a lot 🙏 I will try this out

[drifting coyote]

Thank you! Please feel free to provide your feedback or images you want to share that you make 😊

[drifting coyote]

Lmk if you have any questions

[vivid silo]

Thank you. I will 😉 I need to take some time to digest and try this out over the weekend

[drifting coyote]

You'll love how simple it is

[vivid silo]

Sounds good 🤝👏

[wicked tide]

We appreciate all that you’re doing @drifting coyote

[drifting coyote]

Thank you! I just dropped a bombshell release in the Library! https://discord.com/channels/974519864045756446/1098381588875710484

[wicked tide]

I saw that! Just as I was thanking you too lol

[drifting coyote]

Just dropped the prompt! https://discord.com/channels/974519864045756446/1098381588875710484

@undone trellis wdyt?

[night folio]

anybody here?

[drifting coyote]

Yes

[drifting coyote]

What version ChatGPT are you using? Is it GPT-3.5?

[night folio]

4

Great! Let's start with some questions to gather more information and refine the image prompt further.

What is the subject of the image?
What is the mood or atmosphere you would like to convey?
Are there any specific elements you would like to include or exclude from the image prompt?

J R

1. Manhattan, New York
2. bustling city, full of activity and people going about their business
3. include buildings, cars, people, advertising. exclude anything unrealistic or futuristic.

then it creates a description of the subject not the actual prompt

gpt: Sure, here's a description of Manhattan, New York based on your specifications:

Manhattan is a bustling city that never sleeps, teeming with activity and people going about their daily lives. Tall buildings made of glass and steel loom overhead, casting shadows on the crowded streets below. The sound of car horns and sirens fills the air, as taxis and buses weave in and out of traffic, vying for space on the narrow roads.

People from all walks of life can be seen walking briskly down the sidewalks, dressed in business attire or casual clothes, carrying bags and briefcases, or clutching their phones. The sidewalks are crowded, but people move with a sense of purpose, determined to reach their destinations.

Advertising billboards and signs cover the sides of buildings, their bright colors and flashing lights competing for attention with the crowds of people below. But amidst all the noise and chaos, there is a sense of energy and excitement that permeates the air, a feeling that anything is possible in this city that never sleeps.

[night folio]

no im sorry im using 3.5 thats my problem

[drifting coyote]

Did switching to 4 work for you?

[violet stirrup]

imagine/

[night folio]

yes its working great with gpt4. thank you

[drifting coyote]

🙌 😎

[drifting coyote]

Did you ever get a chancve to take QS OS for a spin?

[wicked tide]

not as much as I'd like! am intimidated i lack the knowledge to utilize properly so observing others who do use it properly to learn

[drifting coyote]

Give it a go! You'll learn far more experimenting with things. Just remember, there wasn't anyone using ChatGPT a few months ago. ⚡ 🧐

[old cedar]

Thanks! I made this. Not quite what I was looking for, but gorgeous!

[drifting coyote]

What were you trying to create?

[old cedar]

'I want it to a photograph of bitcoin taking over the universe. I'd like you to select the camera equipment, lenses, and shot angle. I'd like to include the Earth and the united states.'

https://shareg.pt/NYChed8

[drifting coyote]

Have you used reference images and image weighting?

[drifting coyote]

Just dropped this for the community https://discord.com/channels/974519864045756446/1099407461515022336

[prisma fractal]

How to create images

[drifting coyote]

Do you have MidJourney?

[mint hearth]

Hello are there any script for image prompt template creator?

[drifting coyote]

Can you elaborate? I'm not sure what you mean

[drifting coyote]

It will definitely help you! LMK if you have questions

[mint hearth]

I am trying to generate a batch of football star avatar stickers for my kids...

I am looking for a good prompt to generate some templates so that I could have 2 or 3 different styles of avatar,

Then a good prompt to produce image prompts that I could input into midjourney directly once I provide the bot with templates and list of the football stars

I tried to do this by my own but seemed not doing very well.

I am not sure if this is clear enough

[drifting coyote]

This should do exactly what you want, as long as you have reference images for MidJourney you upload there.

[mint hearth]

I just tried your script, can I DM you? I want to send some screen shots

[shell vapor]

Just tried this and it didn't work off the bat. It interpreted your initial example as the prompt I want

[vale cedar]

Interesting prompt. Quite different approach than the one I used forMidjourney. I have not tested yours. Interestingly - I did not train mine AT ALL on Midjourney syntax or commands - I went 100% descriptive and let the user add whatever they like to the generated description after the fact. Also - mine is primarily photography/cinematic stills - maybe a little trade-off between flexibility and being very good at the specialization?

[drifting coyote]

Certainly

[drifting coyote]

This typically happens when you use it with ChatGPT-3.5. Which version are you using?

[shell vapor]

I used 4 but I just told it that it was using the example and that I’d like to start with a different prompt and it fixed itself

[drifting coyote]

Can you send a screenshot of that interaction? It may be an opportunity to adjust something. Thank you for you feedback 😊

[drifting coyote]

Disclaimer: This response is mostly for other viewers 😎

I appreciate you giving it a go. GPT particularity struggles with highly technical things it doesn't already know, primarily due to initial token input. The challenge with MidJourney is all the required parameters that are fairly rigid, along with the creativity you are asking it to have. Being that my goal was to create a simple, user friendly experience, I inverted the model so that GPT didn't have to consume so much token bandwidth trying to follow the steps and could always put out a structurally correct image prompt. It took many versions to work out the creative side however, not for reasons you think. When V5 was released the MJ model shifted a bit. The best resource I could find regarding this came from @desert sparrow MidJourney Guide that discusses the shift very well. I focused on getting the parts correct so that the user was most likely to get the exact image they were imagining. In addition, I wanted GPT to get the weighting, seed, and image weight factors correct without the user needing to know exactly how they worked. The result wasn't incredibly long and verbose prompts however, it definitely nailed everything I wanted for the average user. The one thing that I will add though, is the stylized and chaos settings are set by default to focus on stable and similar blocks of images so when the user gets their first block of images they require less iterations to get their desired output. I think I will adjust the prompt to add a question that will ask the user of the want a wider range of variety and additional influence from MJ's default style aesthetic.

As always my friend, thank you for your great feedback.

[vale cedar]

This is why I use SudoLang templates instead of 1 shot or few shot examples.

[vale cedar]

I might consider adding weight supports but then you'd need to support multiple syntaxes to get it working with both Midjourney and stable diffusion. Mine currently works with both as-is.

BTW, good job with this. Publishing mine soon! 👏

[drifting coyote]

Looking forward to trying it out!

[drifting coyote]

Yeah, this is a challenging piece for sure

[thin condor]

website of school template

[last heron]

Is it possible to use the page formatting to embed images? Ive seen it done with other ai image generators linked from chatgpt

[copper torrent]

@last heron

[last heron]

?

[drifting coyote]

Yes, you can embed images in your GPT sessions using markdown

[drifting coyote]

Interesting. I'll test and see what my results are

[mint cipher]

I'm a product designer, and I'm searching for methods to make more realistic 3D mockups for my products, such as the one above, by uploading a sample of the product to Midjourney and getting several variants of a more realistic one. What prompt should I use?

[shell vapor]

It won’t be able to take your design and adapt it to the fine detail you want. But it might just wing it and make a good random design by pure chance. Try to use an image of one of the cans in your prompt and then verbalise what you’d like to see. Include the words “product imagery” etc

[drifting coyote]

You can also provide those images as reference however, you'll still need to edit text in another program.

[vale cedar]

@drifting coyote PromptCrafter is here: https://discord.com/channels/974519864045756446/1100695658597462066

[mint cipher]

Much appreciated @shell vapor

[mint cipher]

Thank you @drifting coyote

[regal lark]

How do I pin my favorite prompt library 🥹.

[drifting coyote]

I have yet to see an option to do that. I have reached out to the discord developers to be able to pin messages or posts. Let me know if you find a way.

[ashen heron]

My current work-around is I just "Follow" the thread where the original prompt I like is, and just keep tracking the updated post.

[drifting coyote]

Great idea! Thank you for sharing.

[regal lark]

Great that worked for me

[ashen heron]

Yey! Happy it worked for you! 🙂

[tiny flax]

updated version is in first message ?

[drifting coyote]

No. I will upload the most recent version shortly.

[tiny flax]

thx 🙂

[primal dagger]

@primal dagger

@primal dagger

[drifting coyote]

👇 Here's the Most Recent Version👇 😎

[south goblet]

Awesome work, thank u

[drifting coyote]

Happy to help! LMK know if you have any questions or suggestions for improvements

[tame isle]

this is great, thank you

[drifting coyote]

Thank you for the feedback! I think it works even better now with V5.1 😎

[calm turret]

This is by far one of the best prompts I've worked with. Well done. Suggested next steps for improvement:

-Ask the user about their education level in art concepts

• Educate the user on art concepts (probably just the basics, unless the user wants more???) to better inform/prompt.

Mostly just wanted to post to give MAJOR FREAKING PROPS for

1. Spending the time to research the best way to do this. Anyone that can't see the clear time/effort you put into this is blind

2. Taking the time to share that knowledge with others.

Big props from me. And I'm an autistic cranky old lady. I don't give praise easily 🙂

[red locust]

PLZ GUIDE ME WHERE TO GO FOR CREATE IMAGES STEP BY STEP I AM THANKFUL TO Y0U

[bronze garden]

Was this created with SuperPrompt or Quicksilver OS? @drifting coyote

[drifting eagle]

asdasdsa

[hushed wraith]

hello?

[drifting coyote]

@calm turret Thank you for your kind words and suggestions! 🙏 It was a challenging task but I finally found a stable version by inverting the model and giving GPT the output structure first.

I especially appreciate your comments considering your knowledge base and position 🙂

[drifting coyote]

Do you have MidJourney?

[drifting coyote]

Hello @hushed wraith 🙂

[drifting coyote]

I created this with SuperPrompt 🎯

[bronze garden]

would QS Os be better for making midjourney prompts or would you personally use SP?

[drifting coyote]

I'd use SuperPrompt, as it's specifically designed for building prompts

[deft loom]

@drifting coyote do you have a discord server of your own for your prompts?

[drifting coyote]

I do not but that is now on the list!

[deft loom]

Please send an invite when you do, I am really impressed with your work

[dusty rover]

This works really well for Adobe Firely. went from the right image to the left image.

[drifting coyote]

Thank you for your feedback 🔥

[drifting coyote]

I haven't tried this with FireFly. Now I will! Thank you for sharing

[dusty rover]

in hindsight, should've posted those the other way for a proper "before and after". so it goes. lol.

[drifting coyote]

Either works 😄

[dusty rover]

one good one less good. all works out! lol.

that being said, i do still see value in how that original one looks. just a different style. There is still some kind of beauty to the model guessing? if that makes sense.

obviously I was going for how the better output looks. but I still appreciate the process. ha.

last of my adobe firefly spam, but just showing the overall page with other takes, and the actual prompt. I swapped it over to widescreen/graphic and messed with some of the effects/filters. same theme obviously.

I really hope firefly get's some preset options. via the filters on the right side, you can dial in some standards you usually come back to. I recently threw some feedback at them to start allowing us to save out presets. ha.

[drifting coyote]

Thank you for sharing! This is great for others to see and learn from. Keep creating awesome stuff!

[true veldt]

Yes please, if only for the fact it will be easier to track the latest version of your prompts. They could have a channel to themselves and pinned. 😀

[vale vault]

I need a logo to include the words "دار الكاب " using arabic calligraphy

[drifting coyote]

This can be done in the next version of MidJourney however, for now you would need to add that to your logo after you create it in MidJourney

[mint lynx]

identify this article from "Using inquiry-based writing instruction to develop students’ academic writing skills"
and answer by following

1. The gap found in the articles
   2.The objectives of each research
   3.The method used
   4.The result of the research articles

[mossy quartz]

Are you able to pin the latest version of this prompt to this thread?

[mossy quartz]

From my scrolling this is the latest right? And thank you by the way for sharing these

[drifting coyote]

Yes

[drifting coyote]

It should be pinned in the original post

[little dock]

woah this is awesome Thank you

[drifting coyote]

Thank you 😎 Please come back and share what you make with it 🌟 Also, let me know if you have suggestions for improvement or other feedback.

[tawny whale]

🔥

MJ novice here. Thank you for this!🔥

[drifting coyote]

I hope it helps! I made this specifically to help beginners find MJ much easier to use and get the images you imagine.

[uneven moat]

I'm having problems generating images on midjourney. I paid for the basic subscription just to check it out. When I type in a prompt all it does is keep searching. Ive tried for several days to get nothing. Anyone have any advice?

[drifting coyote]

Are you pasting this prompt into MidJourney?

[hard knot]

is there a setting to turn off the multi weighted prompts?

[hard knot]

v5 and v5.1 get kind of weird when it comes to weighted prompts

[drifting coyote]

You should be able to avoid weighting when it asks you if you want to emphasize anything in your image. Just answer no.

[hard knot]

nice, I will try that. I had a workaround of just asking it to remove numbers and : from the finished prompt

[drifting coyote]

Great work around!

[cobalt kelp]

This is great @drifting coyote - btw are you interested in chatting about some freelance work for some ChatGPT / image generation ideas that are more plugin type but we need help with login on some things etc.?

[drifting coyote]

Sure. Shoot me a dm and we can discuss.

[cobalt kelp]

Sounds good. Will do tonight.

[calm turret]

I do actually have some quality feedback (finally) - I've noticed if you change this from a 1 shot (1 example "excellent prompt") to a 3 shot, the quality output difference is HUGE.

It seems to be less about the commands you present, and more about the examples of quality/accuracy given

[vale vault]

logo for a tropical guesthouse named Dar ElCap

[odd heart]

don't forget the version parameter --v 5.1 to get the best results, as far as i know the default model isn't the latest and best one available, you have to specify that; how would we add this to the prompt?

"/imagine: a stunning photograph of a boy sitting on a remote Maui beach at sunset::3 in a golden hour setting::3, featuring vibrant colors reflecting off the water, palm trees framing the scene, and footprints in the sand::2, inspired by Peter Lik and Ansel Adams, using a low camera position and slightly tilted perspective, and warm hues and long shadows from the golden hour sun::2. --ar 16:9 --q 2 --chaos 0 --stop 100 --s 100 --v 5.1"
... omitted for brevity
8. ...
•    Version: Use "--v" followed by a number to specify the version of the AI model you want to use for generating the image. For example, "--v 5.1" for the latest version as of 2023-05-22.

[mossy quartz]

currently the default model for MJ is version 5.1, if you don't specify a 'v' parameter at the end it will default to 5.1

[odd heart]

Oh OK that makes sense. My bad.

[mossy quartz]

no worries! just figured id let you know

[calm turret]

Used your prompt, I then asked GPT to write a creative example of a multimodal AI (referencing many models GPT knows) that had together achieved "something that we understood to be consciousness". It gave me this midjourney prompt, and the screenshot is the midjourney output in my personal server.

Both the prompt and the image are beautiful, so I thought you'd appreciate seeing it OP

[boreal merlin]

5.1 has been the default since 5/4 🙏

[odd heart]

Good to know! In my defense, inclusion of version parameter would be a best practice for reproducibility

[boreal merlin]

Haha absolutely a good point!

[drifting coyote]

I'm listening

[drifting coyote]

I'll be updating soon 😌

[drifting coyote]

This is a great example 😳😊

[calm turret]

Did I forget to give this lol? First is adding more quality example prompts. Exponential creative output improvements - I use a few from the highest rated public prompt section of MJ prompt.

It seems like 3 is the sweet spot, but obviously 5 shot is great too

[drifting coyote]

You did 😁

[drifting coyote]

I'll definitely modify accordingly

[calm turret]

Let me not be lazy and actually give you the examples I'm using quick

As you can see, I also tried to vary the AR of the examples I gave and the TYPE of artistic output. I think honestly the "examples" section seems to help the MOST (which makes sense - in research we see 25 shot > 1 shot)

I was super lazy about the examples though. I just grabbed 3 additional quick ones. If you actually put time into providing QUALITY varying examples???
dramatic Italian hand gestures of perfection 🤌

[bright cliff]

Thanks a lot. I love it. I use this to generate photorealistic images using a collection of good prompts. Will adapt it a little, but thanks for the having taken the time to present it out so structured 💕

[calm turret]

No problem at all friend, but definitely know this is @drifting coyote's work and prompt, I was just providing feedback and proposing a VERY small edit to a quality prompt 😄 He's the true prompt artist! ❤️

[drifting coyote]

🙏😊

[violet sun]

wow

[finite badge]

are you turkish @drifting coyote

[cinder geyser]

Is this for the API? The website has a fairly small character limit that makes it hard to outline realism, etc.

[foggy ice]

Quicksilver, MadameArchitect, is it possible to create a specific kind of artwork based off a Action in photoshop using a .Atn
file?

[drifting coyote]

🙂

[drifting coyote]

I am Anglo-saxon my friend.

[drifting coyote]

No, this is for front end use with MJ...however, you can use it with the API

[drifting coyote]

Yes. Do you know how to set up web based .Atn files or functions?

[foggy ice]

No, I wish. but i am sure i can learn.

[drifting coyote]

Look into how to set up macro functions that are web based

[foggy ice]

I have the action that automates it, but i don't think I can incorporate it into chatgpt

[upper bay]

meta MJ

[drifting coyote]

Nope, not yet

[drifting coyote]

Hello @upper bay

[untold jackal]

@drifting coyote will there be a Update from this prompt to get Stable Diffusion prompts?

[drifting coyote]

Yes! I am launching a discord server and Git this next week!

[dull juniper]

@drifting coyote I'm curious if you've been working on anything related to prompt generation for Adobe Firefly. Looking forward to the discord server and git btw.

[drifting coyote]

I haven't done much with firefly however, with some tweaks, MJ works. I will be addressing what to focus on, given that there are a lot of different platforms trying to own the space. I am excited about firefly for sure

[dull juniper]

I see.

I've been experimenting with Firefly almost daily for a month now tbh. It's still in beta and there's a lot of room for improvement, but considering it's Adobe, I expect them to put out a quality product by the end of it.

[drifting coyote]

I'm am hopeful however, sometimes the larger companies lean towards commercial benefits more than creative or cutting edge creative tools that redefine artistic palettes

[dull juniper]

That is true

Well, I wouldn't lose hope, tho
We will probably have open source tools for that or new startups

Unless regulators decide to kill everything

[drifting coyote]

Exactly

[dull juniper]

A huge problem everywhere tbh

With many officials not knowing how technologies work and letting companies lobby them

[drifting coyote]

Unfortunately

[hard knot]

This still a thing?

[lucid pendant]

provide us with latest prompt! 😄

[ebon musk]

Hi hi, a new release of mj was released today 😀 some cool new features. Perhaps bake them into your super super prompt. Thanks for your inspiring work.

[drifting coyote]

I am updating for sure 😎

[ebon musk]

The 'weird' option 😄👠

[tepid quartz]

@drifting coyote any news? 🙂

[elder river]

i did a brief introduction for anyone that is curious on utilizing META mj for your midjourney photos. checkout GPTtogether.

[foggy ice]

Got this using this awesome prompt creator. Kobe

[drifting coyote]

Sick 🔥

[foggy ice]

Thank You

[bitter pond]

where is the most recent Version?

[storm gulch]

This is so good, thank you very much!!!

[drifting coyote]

😎 Glad it is useful to you.

[drifting coyote]

😎

[rapid spade]

@drifting coyote Awesome prompts! Thank you for sharing al this. Did you make a Discord server? Where might I get an invite and/ or download latest superprompt version?

[broken steeple]

where is the prompt -_-

[clear mirage]

Hi @drifting coyote - Can I get an invite to your server?

[stray stone]

Where is the prompt?

[stray stone]

Please can you share with me the prompt

[patent pasture]

Noce

[coral vessel]

I thought it creates the image....

oh wait it can and the images are just like midjourney

[hidden sorrel]

I’m hiring video editors, graphic designers, and creative generalists.

I’m looking for people who can take strong references, turn them into usable concepts, understand content structure, and execute fast with clear communication.

Pay

Video editors: $60–$120 per video (based on complexity)

Creative generalists (2–3 yrs): $150–$200 per video

Graphic designers: Flexible, based on portfolio and output

You’ll work on high-quality content for founders, YC startups, and international clients.
I’m building a serious creative team and scaling quickly good people will grow fast here.

[rustic elm]

@drifting coyote I see your still here indulging in chatgpt shenanigans, haha. And buy the way people, this guys creations are nothing short of amazing🔥🔥