OpenAI launches ChatGPT Images 2.5, which it says cuts image generation latency by up to 50% vs. Images 2.0, and adds a Sketch feature for drawing in ChatGPT (Ina Fried/Axios)
ThinkingNews Desk · how this was written
OpenAI has released ChatGPT Images 2.5, a new image-generation model that reduces latency by up to 50% compared with Images 2.0. The update adds a Sketch feature that turns users’ rough drawings into detailed AI-generated images.
Written from all 3 reports below, not from any single one.
How it was reported
- Hacker News·ChatGPT Images 2.5
- TechMeme·OpenAI launches ChatGPT Images 2.5, which it says cuts image generation latency by up to 50% vs. Images 2.0, and adds a Sketch feature for drawing in ChatGPT (Ina Fried/Axios)
ChatGPT Images 2.5 cuts image-generation latency by up to 50% versus Images 2.0 and introduces a Sketch tool for direct drawing in ChatGPT. OpenAI reports an internal model, described as far more capable than GPT-6 Astra, solved the Navier-Stokes Millennium Prize problem using 10 000 concurrent agents running for 88 hours. The company cannot rule out that de-identified data from mathematicians Tristan Buckmaster and Levent Alpöge contributed to the model’s development.
- The Verge·ChatGPT Sketch turns your bad drawings into detailed AI images
OpenAI’s ChatGPT Images 2.5 adds a Sketch feature that lets users draw a doodle inside ChatGPT by typing @Sketch, then instruct ChatGPT to generate an image from that sketch. In a test, a rough mouse-drawn cat was converted into a realistic photo per the user’s prompt. The tool also allows comments to be added directly to specific parts of the generated image.
Related stories
- OpenAI upgrades ChatGPT with interactive learning tools as lawsuits and Pentagon backlash mount5 outlets
- "Chat is dead": OpenAI preps overhaul of ChatGPT4 outlets
- ChatGPT's new default model is more factual and better at personalization4 outlets
- OpenAI rolls out GPT-5.6 after government green light — and announces ‘ChatGPT Work’6 outlets