Introducing ChatGPT Images 2.0 A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence. Video made with ChatGPT Images

ChatGPT Images 2.0 is a step change in detailed instruction following, placing and relating objects accurately, and rendering dense text, with the ability to generate across aspect ratios. It’s also accurate across languages and uses its expanded visual and world knowledge to fill in the gaps for you, so you get smarter images with less prompting.

Greater Precision and Control ChatGPT Images 2.0 can conceptualize more sophisticated images, and then actually bring that vision to life effectively. It’s able to follow instructions, preserve requested details, and render the fine-grained elements that often break image models: small text, iconography, UI elements, dense compositions, and subtle stylistic constraints, all at up to 2K resolution. Zoom in on the rice…

Stronger Across Languages ChatGPT Images 2.0 can produce images with non-English text that’s not only rendered correctly but with language that flows coherently. This makes the model more globally useful and helps people create visuals that work in the languages they actually use.

Stylistic Sophistication and Photo Realism ChatGPT Images 2.0 is better able to capture the defining characteristics of photos, as well as cinematic stills, pixel art, manga, and other distinctive visual languages, with greater consistency in texture, lighting, composition, and fine detail. This is especially useful for game prototyping, storyboarding, marketing creative, and creating assets in a particular medium or genre.

Flexible Aspect Ratios ChatGPT Images 2.0 supports aspect ratios as wide as 3:1 and as tall as 1:3. It can generate outputs that are ready to fit the formats you need, from wide banners and presentation slides to posters and social graphics.

A Visual Thought Partner ChatGPT Images 2.0 is our first image model with thinking capabilities. When a thinking model is selected in ChatGPT, Images 2.0 can search the web for real-time information, create multiple distinct images from one prompt, double-check its own outputs, and even create functional QR codes. This allows it to take on more of the heavy lifting between idea and image, especially when accuracy, up-to-date information, consistency, and visual cohesion matter most.

Real-World Intelligence ChatGPT Images 2.0 has an updated knowledge cutoff of December 2025 and intelligence that allows it to expertly handle tasks end-to-end, from copywriting to analysis to design composition.

ChatGPT Images 2.0 is available starting today to all ChatGPT and Codex users. Images with thinking are available to ChatGPT Plus, Pro, and Business users (Enterprise soon). On mobile, make sure you update to the latest version of the app. The underlying model, gpt-image-2, is available in the API.

🤣🤣

Tested it. It’s easier to recognize the person in the one drawn with colored pencils.

The model is so good

“A visual thought partner” OpenAI always pivots to collaboration when it’s time to advertise, but lies about deprecation timelines once you actually integrate their models.

We officially entered an era where AI images are indistinguishable from real images especially excited about proper text generation on images

This being fully generated with ChatGPT Images 2.0 is pretty impressive. Two step process. One for the portrait, one for new magazine front.

nerfed by May

well my previous meme got stolen so here is the new one 👀

ran one prompt for a full brand identity. wasn't expecting this

Congrats on the fantastic progress! As a hand surgeon, I run my own informal evals on image gen models. 1. The hand x-ray is the best AI-generated hand x-ray I've ever seen. The carpal bone anatomy is particularly complex and well done. The distal radius has an odd feature, but I'm not sure I would immediately recognize this as AI generated. 2. The carpal tunnel informational handout is really polished, but unfortunately the anatomy is far too inaccurate to make it usable. The gap between "looks right" and "is right" is rapidly closing for medical images. Congrats!

Gpt image v1.5 vs v2 Same prompt, difference is honestly impressive

41% chance OpenAI announces a head-mounted display this year.

well done. something like this is pretty much impossible with other models.

I first gave it some context for image content: “Based on your memory and our past conversations, create my bio and brand story” 1) “Create an image of a two-page business magazine spread with my story. No photo of me but use graphic design elements commonly found in magazine articles” 2) “Using what you know about me, create an image of Wikipedia page about me” 3) “Using what you know about my skills, values, traits and expertise, create an image of an app UI that delivers results based on the value I can create. Use attached photo of me to add my image to the design” 4) “Based on my personal mission and values, create an image of a text message exchange on a smartphone between me and my future self giving me advice” 5) “Inspired by my brand story, Create an sci fi action movie poster about futuristic business competition tied to time travel into 1800-1900’s business endeavors. Include elements commonly used on movie posters including punchy review quotes with film award wreaths, cast names, “filmed in ____” etc”

PLEASE DONT NERF IT PLEASE DONT NERF IT PLEASE DONT NERF IT PLEASE DONT NERF IT

My very first GPT-2 creation... and it already slaps 🔥

I think Elon broke media play back today on purpose due to your announcement 😂

#keep4o

Tried it out, the manga is actually pretty good for character consistency. RIP OpenAI servers however as they slowly get more and more people logging in to try it out.

Almost every prompt: “We’re so sorry, but the prompt may violate our guardrails concerning similarity to third-party content. If you think we got it wrong, please retry or edit your prompt. I wasn’t able to generate the image due to an error on my side.”

Very soon, no would would be able to tell the difference between real Vs AI generated slop

WOW, NO SIGNLE MISTAKE

#shutterstock entire business is images. OpenAi Images 2.0 realism is unparalleled.

wow the images generated by this is now almost undetectable as AI i hope you have better safeguards in place

So much detail, small text, no errors, knowledge of ChatGPT. @OpenAI cooked on this one.

@sama lets make it official!

This was one of the best promo videos for an image generator I've actually seen. The quality and the precision adult, but the creativity is, next level. Someone deserves a raise.

gpt images v2 hit a new level for brand kits

awesome for brand kits

$CAM is here to stay.

So it's real, folks. Thank you for confirming @OpenAI . 🦎

Impressive and amazing!