OpenAI’s Images 2.0 Drop
- OpenAI announced ChatGPT Images 2.0 for advanced image generation and editing in a widely shared X post. (x.com) - The post logged tens of thousands of likes and millions of views, signaling heavy user engagement around the launch. (x.com) - The rollout is already part of broader social conversation about AI art tools joining traditional studio and timelapse workflows. ( )
OpenAI said on April 21 that ChatGPT Images 2.0 is rolling out inside ChatGPT with a new image model for generation and editing. (openai.com) OpenAI’s product post said the update improves text rendering, multilingual output, and “visual reasoning,” the system’s ability to keep more parts of a prompt consistent inside one image. OpenAI’s release notes said ChatGPT Images 2.0 is available on all ChatGPT plans. (openai.com; help.openai.com) The Help Center said users can create images from prompts, upload an existing image, and edit selected parts of an image with a built-in selection tool. OpenAI said image generation can take up to two minutes, depending on prompt complexity. (help.openai.com; help.openai.com) Image generators turn text instructions into pictures, and editing tools let users change one region without remaking the whole frame. OpenAI has been moving that workflow from separate image products into ChatGPT itself since it introduced 4o image generation in March 2025 and a redesigned ChatGPT Images experience in December 2025. (openai.com; openai.com) The April 2026 release extends that push by making image creation part of the main ChatGPT interface rather than a side tool. OpenAI’s Help Center also said “images with thinking,” a mode that adds extra reasoning before image output, is available on Plus, Pro, and Business and is “coming to Enterprise and Edu soon.” (help.openai.com) OpenAI’s earlier image releases focused on practical tasks such as accurate text in posters, product mockups, transparent backgrounds, and iterative edits to the same composition. The company’s March 2025 API launch for `gpt-image-1` also said generated images include Coalition for Content Provenance and Authenticity, or C2PA, metadata, a standard for labeling media provenance. (openai.com; openai.com) OpenAI has also published safety material for its image systems, including a GPT-4o image generation system card addendum that said the model can create photorealistic output and transform uploaded images. That document said OpenAI evaluated risks including harmful image generation as image tools became more capable. (openai.com) The company’s latest post puts the emphasis back on everyday use: make an image, pick a section, and ask for a change in plain language. After a year of folding image tools into GPT-4o and then ChatGPT, OpenAI is now treating image editing as a default chat feature, not a separate app. (help.openai.com; openai.com)
