OpenAI has launched the ‘thinking’ image generator ChatGPT Images 2.0 — a “cutting-edge model capable of tackling complex visual tasks and creating precise, ready-to-use works.”
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 pic.twitter.com/3aWfXakrcR
— OpenAI (@OpenAI) April 21, 2026
The company highlighted a “qualitative leap” in following instructions, precise placement and proportion of objects, as well as in visualizing dense text.
The model confidently operates in multiple languages and autonomously fills in gaps in requests, relying on visual and general context. As a result, users achieve the desired outcome with fewer clarifications.
Precision and Control
ChatGPT Images 2.0 handles complex concepts and accurately brings them to life visually.
The model follows instructions, retains specified details, and displays fine elements with a resolution of up to 2K.
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… pic.twitter.com/n29165pV9Q
— OpenAI (@OpenAI) April 21, 2026
Working with Styles
ChatGPT Images 2.0 more accurately conveys the distinctive features of photographs, cinematic frames, pixel art, manga, and other visual styles. LLM ensures a high degree of consistency in textures, lighting, composition, and fine details.
This level of precision can be beneficial in creating game prototypes, developing storyboards, preparing marketing materials, and creating works in a specific media format or genre.
Capable of Thinking
ChatGPT Images 2.0 is OpenAI’s first image model capable of reasoning before generation.
In conjunction with ChatGPT, the model can search for information online in real-time, create multiple variations from a single prompt, verify results, and generate functional QR codes.
“This allows the model to take on much of the heavy lifting between idea and image, especially when accuracy, information relevance, consistency, and visual integrity are paramount,” OpenAI claims.
The model supports aspect ratios from 3:1 in width to 1:3 in height. It is available to ChatGPT and Codex users.
The Images with thinking feature is included in the ChatGPT Plus, Pro, and Business plans.
In April, OpenAI granted limited access to its new AI model GPT-5.4-Cyber to select users.