The Official 4o and Dall-E image Megathread

[jeffvpace]

DALL-E 3 max prompt size is 4000 characters. DALL-E 2 max prompt size is 1000 characters.

[phyde1001]

Maybe there was an update since the cookbook?

• prompt (str): A text description of the desired image(s). The maximum length is 1000 characters. Required field.

What's new with DALL·E 3? | OpenAI Cookbook

@PaulBellow

You can find more in the docs for image endpoint or the new DALLE3 cookbook page

You never know who you can trust these days

A text description of the desired image(s). The maximum length is 32000 characters for gpt-image-1, 1000 characters for dall-e-2 and 4000 characters for dall-e-3.

https://platform.openai.com/docs/api-reference/images/create

[jeffvpace]

You never know who you can trust these days

No biggie - OpenAI is notorious for not updating docs…

However, @PTXeno’s prompt does exceed 4000 chars. I use the API (not ChatGPT or Playground) so I have created UI’s for all of the models we use and I could not use the prompt - ERROR 400.

[phyde1001]

I use both and guessing it balks at some of those characters too…

° ~ etc

[jeffvpace]

Yup, there are special (non-ASCII) characters in there. I suppose ChatGPT is more foregiving for stuff like that - it probably did truncate.

[windysoliloquy]

afternoon boys!

i was pretty impressed with my first gen spiritual good guys….
i’ll edit later to not interrupt

mmmmmm

[jeffvpace]

“Yarn-Mode” good for mummies

[Tina_ChiKa]

Hybrids:

Your master prompt coupled with my approach

[jeffvpace]

Muy bueno. Sehr gut. Très bien

[Tina_ChiKa]

Thanks. Danke. ありがとう

(attachments)

[PTXeno]

Color Palette::Reflective viscous black with mirror-like sheen | Scene::Radiant kitsune goddess, half-fox woman; near-profile face upturned, eyes half-closed in ecstasy, torso twist, silken white fur | Camera::Low Angle | Lens::100 mm-macro | Lighting::Baroque Candle chiaroscuro | Grid::Golden Triangle | Aura::Pure white glow | Sakura petals on wind | Emotion::Euphoria | Atmosphere::Dark extreme ultradramatic | Medium::digital oil painterly strokes | Style::D&D 5e—Tyler J

-

With Bing

[Bleatlock]

[windysoliloquy]

[creativeflightsstudi]

Playing with an idea I have about a watercolor painting coming to life. Again pushing DALL-E to render two styles…watercolor and photorealistic.

There is still a bit of the Van Gogh styling artifacting but I like it.

ChatGPT - DALL-E 5

[phyde1001]

Inspirare

[windysoliloquy]

I could be fine tuning these for weeks…
probably sidestep to something less dramatic soon.

[creativeflightsstudi]

@phyde1001 That is awesome! I love it!!

[PaulBellow]

[chisanaminamoto]

[windysoliloquy]

I’m perpetually minded for the future, it’s in my nature so much that I often miss the right now.

And right now I’m trying to focus on things that I can add to my image library, that one day can turn into an open world game where everything is rendered as realistic as possible out of yarn.

I figure by the time I’m done adding creatures and things it should be fairly easy to feed it all into an Ai workflow and turn it into a game.

Would you guys wanna help me with some NPC ideas? I’m kinda stuck on things from my own gaming experiences which is currently twitting out on evolving the deku tree into something not copywritten…

And maybe perhaps, have different communities centered around different talking trees: