On November 4 OpenAI launched Images 2.6 across all ChatGPT and Codex tiers. Users can turn a prompt or image into a looping GIF and save reference images for reuse. The API gets two new models, GPT-Image-2.6 Flare and Sunburst, which generate images 50% faster than before; Flare can produce an image in about six seconds at low quality. The API also adds image markup edits, SVG export and a new complexity parameter.
Key Takeaways
- ✓Available today in ChatGPT and Codex on all subscription tiers across web, desktop and mobile
- ✓New GIF creation with adjustable speed and per-frame editing, even from a simple sketch
- ✓New API models: GPT-Image-2.6 Flare (as fast as about six seconds at low quality) and Sunburst (better consistency across edits at medium or high quality)
- ✓The API adds image markup edits, SVG export and a complexity parameter that controls how much the model plans
- ✓OpenAI says people create about 4 billion images per week across ChatGPT, Codex and the API
Key Decision Metrics at a Glance
Turn your technical choice into a development budget
Compare 40 dev plans & simulate token costs vs $20/mo subscriptions
Project Links & Resources
Direct AccessIn-Depth Technical Analysis
Background and the Industry Pain Point
Image generation is one of ChatGPT's most used features: OpenAI says people create about 4 billion images per week across ChatGPT, Codex and the API, more than half a billion a day. Marketing, game and app teams still run into two problems, slow generation and characters, products or brand details drifting across edit rounds, and static images don't cover the demand for animated content on social media.
Architecture Highlights and Mechanism
Images 2.6 generates images 50% faster than before. The API offers two models: GPT-Image-2.6 Flare is tuned for speed, producing an image in as little as six seconds at low quality, with better layout for posters and product shots; GPT-Image-2.6 Sunburst is tuned for quality and keeps details more consistent across edits at medium or high quality. ChatGPT adds GIF creation with adjustable speed and per-frame editing, plus reference images you can save and reuse. The API adds image markup edits, SVG export and a complexity parameter that controls how much the model plans, set separately from image quality.
Benchmarks and Measured Results
OpenAI's announcement publishes no benchmark scores beyond the 50% speedup and Flare's roughly six-second generation. Third-party DesignArena posted on X that GPT Image 2.6 Sunburst ranks first on its Image Arena, ahead of Image 2.5 Sunburst; that is a third-party leaderboard claim, not an OpenAI figure.
Getting Started for Developers
ChatGPT and Codex users can use it today, and image edits in Codex now happen directly in the overlay. Developers can call GPT-Image-2.6 Flare and Sunburst in the API; pricing is on OpenAI's API pricing page. OpenAI recommends starting with low complexity for most edits and trying medium or high for scenes with many elements or constraints; pick Sunburst for multi-round edits that need high consistency. Safety evaluations are in the Images 2.6 system card.
Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.