OpenAI ships ChatGPT Images 2.5 with precise local edits, Sketch input, and 50% faster API model
APPSO · wechat · 2026-09-09
OpenAI released ChatGPT Images 2.5, noting users now generate over 3 billion images weekly via ChatGPT Images and the GPT Image API. The new model focuses on detail quality, speed, and consistency across multi-turn edits, and adds Sketch hand-drawn references, templates, and comment-based image editing.
- Precise editing: the headline improvement — changing a single element (product, background, text) no longer alters faces or overall style; tested cases show stable identity preservation and support for recursive visual nesting prompts
- Sketch & workflow: draw rough layouts directly in ChatGPT as visual references (works for GIFs too), start from templates, and annotate images to specify edits
- API duo: GPT Image 2.5 Flare is the default with 50% lower latency vs GPT Image 2; GPT Image 2.5 Sunburst targets premium creative workflows like ad assets and product photography
- Caveats: noise artifacts remain a known weakness, and some official examples show detail flaws. Available now to ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web
Related event: OpenAI launches ChatGPT Images 2.5 with Flare and Sunburst API models(43 posts)→
More from Multimodal
- Fish Audio ships open-weight Drama 3 TTS; creator shows turning old posts into narrated videos — alex_verem · 2026-10-27
- Redditor crafts photorealistic 'morning bus ride' short film entirely in ComfyUI — no3us · 2026-10-27
- New blur/anti-blur slider LoRA for MOSS lets you dial background sharpness with weights — NoMouse9610 · 2026-10-27
- PixGR: Pixel-space generative decoding for fast feed-forward 3D reconstruction — kwangmoo_yi · 2026-10-27
- Krea-2 Turbo + ComfyUI for product ads: prompts and results shared — comfy-king-999 · 2026-10-27
- High-res tiled Marigold V2 depth maps arrive in ComfyUI with See Through model — AntonObukhov1 · 2026-10-27