OpenAI has released ChatGPT Images 2.6, an update to its image generation system that adds GIF creation and reference image support, and announced two new image models for its API that it says generate images 50% faster.
Product update · OpenAI
ChatGPT Images 2.6 adds animated GIFs, reference images, and faster API models
The update extends ChatGPT's image tools in two directions for consumers, while OpenAI says two forthcoming API models generate images 50% faster. Neither the model names nor an API release date have been announced.
50%
Faster image generation claimed for the two new API models
2
New image models coming to the API, not yet named
2
New ChatGPT features: GIF creation and reference image input
Claimed generation speed, relative to current
Illustrative scale: each block is 10% of current speed. The 50% gain is OpenAI's figure; no absolute timings were given.
New API models
150% (claimed)
What the update enables
Animated output: GIFs, not only static images
Reference images to guide style and subject
Faster generation that could cut interactive wait times
Possibly lower cost for high-volume requests
What remains unverified
No pricing or benchmark methodology provided
No quality comparison with earlier versions
No independent testing available at reporting time
No general availability date for the API models
What will decide the update's impact
Whether it changes creative and developer workflows, or is an incremental step, depends on three tests.
Frame consistency
Does the subject hold steady across animated GIF frames?
Reference fidelity
How faithfully do outputs follow the supplied reference image?
Speed under load
Does the 50% gain hold up under real-world request volume?
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...