Understand this story
On November 5, 2026, OpenAI released Images 2.6, launching a GIF generation feature for ChatGPT: users can enter text or upload an image to generate a looping GIF, and can adjust playback speed and view and edit individual frames. The new model is 50% faster overall, and the API offers two versions, GPT-Image-2.6 Flare and Sunburst, while also opening up SVG vector export and an image annotation editing interface; image output pricing remains the same as the previous generation. On November 6, Tibo said that ChatGPT and the API were already generating more than 4 billion images per week, up from more than 3 billion 8 weeks earlier. On November 7, Artificial Analysis's hands-on testing showed that GPT Image 2.6 took the top two spots on both the AA-Image-T2I v2.1 and AA-Image-Editing v2.1 leaderboards, with Flare's median text-to-image time dropping from about 55 seconds to 30 seconds.
AI-generated from public reports
Latest development · Report
GPT Image 2.6 Tops the Artificial Analysis Image Leaderboard, Flare Nearly Doubles in SpeedDevelopments
3 confirmed developments
- Source:Artificial Analysis
Tibo says ChatGPT and API generate over 4 billion images per week
Source:TiboOpenAI launches GIF generation feature for ChatGPT and introduces Images 2.6
Source:Phoenix Technology
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 7
Artificial AnalysisSelected
GPT Image 2.6 Tops the Artificial Analysis Image Leaderboard, Flare Nearly Doubles in Speed
Artificial Analysis's hands-on testing shows that GPT Image 2.6, released by OpenAI on November 4, swept the top two spots on both the AA-Image-T2I v2.1 and AA-Image-Editing v2.1 leaderboards. GPT Image 2.6 Flare's median text-to-image generation time dropped from about 55 seconds to 30 seconds, and editing from about 58 seconds to 32 seconds, with a slight improvement in image quality; GPT Image 2.6 Sunburst takes about 2 minutes 15 seconds per image and claims first place, with the biggest gains in reasoning, text rendering, lighting, and materials. Both models support text-to-image and image editing, and add animated GIF and SVG output; API pricing is $30 per 1 million image output tokens, roughly $0.21 per 1024x1024 image.
falRelated discussion
fal joined this discussion
This post contributes to heat; it is not a formal report or confirmed fact.
Nov 6
China.com Technology
OpenAI Launches Image 2.6 Image Model, PixPix First to Integrate
OpenAI has launched the Image 2.6 image generation model, which improves image generation speed while adding animated GIF creation capabilities. It supports creating GIFs through text descriptions or uploading photos, and allows adjusting playback speed, as well as viewing and editing individual frames. This update also introduces two API-oriented versions, GPT-Image-2.6 Flare and Sunburst. PixPix is the first to integrate this model, and users can experience it on its platform via the entry point pixpix.art.
China.com TechnologyRelated discussion
China.com Technology joined this discussion
This post contributes to heat; it is not a formal report or confirmed fact.
sspai
Morning Brief: General Office of the State Council Releases 2027 Partial Holiday Schedule, and More
The General Office of the State Council released the 2027 partial holiday schedule. The Spring Festival holiday will run from February 5 to 13, nine consecutive days off. Only the Spring Festival, Labor Day, and National Day involve adjusted rest days this year, with a total of 5 weekend make-up workdays. OpenAI launched the image generation model Images 2.6, bringing native GIF animation generation to ChatGPT. The company says image generation speed has increased by about 50% compared with before, and it simultaneously launched GPT-Image-2.6 Flare and Sunburst in the API. Moonshot AI released the Kimi K3.1 Pro weights on Hugging Face. It uses a MoE architecture, has 2.8 trillion total parameters, and supports a context of about 1 million tokens.
Tibo
Tibo: Weekly image generation across ChatGPT and the API has grown from 3 billion to 4 billion
Tibo said that ChatGPT and the API now generate more than 4B images per week, up from more than 3B per week eight weeks ago. The figures were released after OpenAI announced ChatGPT Images 2.6, which adds GIF and reference image features, and two models with 50% faster image generation are about to enter the API.
Thurrott
OpenAI launches Images 2.6 in ChatGPT and Codex, supporting GIF generation from a single prompt
OpenAI has launched the Images 2.6 model in ChatGPT and Codex, with generation speed increased by 50% compared with before, and support for generating animated GIFs from a single prompt or sketch. The API offers two versions: 2.6 Flare produces an image in 6 seconds at low-quality settings, while 2.6 Sunburst is better suited to professional visual assets. OpenAI says ChatGPT, Codex and the API generate more than 4 million images per week, and all generated images carry an invisible SynthID watermark.
Nov 5
36Kr · AI
OpenAI Updates Images 2.6: ChatGPT Now Supports One-Sentence GIF Generation, API Opens SVG Export
OpenAI has updated its image model to Images 2.6. In ChatGPT, users can generate looping GIFs by typing a single sentence or uploading a photo, and can adjust playback speed as well as view and edit individual frames. The new model is 50% faster overall. On the API side, two versions have been released: GPT-Image-2.6 Flare and Sunburst. At low quality, Flare can produce an image in as fast as about 6 seconds. The update also adds developer-facing SVG vector export, an image annotation and editing interface, and a complexity parameter. Image output for both Flare and Sunburst is priced at $30 per million tokens, unchanged from the previous generation.
@mark_kRelated discussion
@mark_k joined this discussion
This post contributes to heat; it is not a formal report or confirmed fact.
Numerama
ChatGPT launches GIF generation with Images 2.6
On November 4, 2026, OpenAI released the image generator Images 2.6; ChatGPT can generate looping GIFs based on text descriptions or uploaded static images, and supports continued modifications in the conversation before downloading. The author tested generating an animation from an image of Tony Tony Chopper; the result was less natural than the pink dancing cat, and also slowed down an existing GIF by providing a link. The feature is available to all ChatGPT and Codex plans worldwide, but GIF generation is still being gradually rolled out and may not yet be available on some accounts; OpenAI says this version also increases image generation speed by 50% and supports calling named reference photos via @ mentions.
Testing Catalog
November 5 AI Daily Brief: OpenAI Releases Images 2.6 and GPT-6.2
OpenAI launched Images 2.6, which supports GIFs and reference images, on ChatGPT and Codex, and introduced the Flare and Sunburst models in its API; GPT-6.2 Astra and Sol have appeared in Google Cloud quotas, and The Information says they could be released as soon as Friday. Kimi K3.1 Pro with open weights is now available on Hugging Face, Claude Code's background sub-agent status updates use about 99% fewer tokens, and GitHub Copilot has entered public beta as a built-in coding assistant in Xcode.
Alibaba Cloud Developer Community AI
ChatGPT Images 2.6 Released, Supports Direct GIF Generation
OpenAI has released ChatGPT Images 2.6, adding the ability to directly generate GIF animations, with image generation speed increased by 50% and better consistency across multi-turn edits. The author tested 7 cases, covering memes, real-person photos, sketch animations, and WeChat chat logs, and provides the complete prompts. After a GIF is generated, playback speed can be adjusted across 8 levels from 0.25x to 2x, the playback range can be trimmed, and users can switch to a sprite sheet to edit dialogue frame by frame; however, real-person fidelity is mediocre, most GIFs are only three to four hundred pixels wide, and moderation occasionally produces false positives.
English text is being prepared.Related discussion
English text is being prepared. joined this discussion
This post contributes to heat; it is not a formal report or confirmed fact.
heise online
OpenAI releases ChatGPT Images 2.6, supporting GIF and SVG vector graphic generation
OpenAI has released the image model Images 2.6; for the first time, ChatGPT can generate animated GIFs—users only need to describe the scene and motion, or upload a photo or sketch to make it move, and the finished product can also have its playback speed adjusted and be modified with text instructions. OpenAI says the new model's image generation speed has increased by about 50%, and it adds savable and reusable reference images, editable SVG vector graphics on the API side, and local box-selection editing; the two API variants Flare and Sunburst have been upgraded in tandem, and the price per million image tokens remains unchanged at $30. Images 2.6 is now available on ChatGPT, all Codex tiers, and the API; the GIF feature is being rolled out in batches, and OpenAI says about 4 billion images are currently generated each week through these three entry points.
Phoenix Technology
OpenAI Updates Images 2.6: ChatGPT Supports One-Sentence GIF Generation, API Opens SVG Export
OpenAI has released Images 2.6. In ChatGPT, users can generate looping GIFs by entering text or uploading an image, and can adjust playback speed and view and edit individual frames. The new model is 50% faster overall. On the API side, two versions are offered, GPT-Image-2.6 Flare and Sunburst; Flare can produce an image in as fast as about 6 seconds at low quality. It also opens SVG vector export and image annotation editing interfaces to developers. Image output pricing for both Flare and Sunburst is $30 per million tokens, unchanged from the previous generation.
The Decoder
ChatGPT now supports generating animated GIFs based on the OpenAI Images 2.6 model
ChatGPT now supports generating animated GIFs, a feature based on OpenAI's new Images 2.6 model that is rolling out to all ChatGPT and Codex users on web, desktop, and mobile. Users can generate GIFs using text prompts or uploaded photos, and adjust animation speed or edit individual frames; OpenAI says the model maintains consistent character appearance across frames. For developers, OpenAI also released the API models GPT-Image-2.6 "Flare" and "Sunburst," focused respectively on efficiency and higher quality, and said users currently generate more than 4 billion images per week through ChatGPT and the API.
Unite.AI
ChatGPT Launches GIF Animation Generation and Editing Features with Images 2.6
OpenAI has launched native GIF animation generation in ChatGPT, which can turn text prompts or uploaded photos into GIFs, and supports adjusting playback speed, modifying individual frames, and frame-by-frame viewing. The update rolls out with Images 2.6 to all ChatGPT and Codex users on web, desktop, and mobile, while the API side adds two new models, GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst, focusing respectively on generation speed and output quality. OpenAI says the total number of images generated weekly by ChatGPT and the API has exceeded 4 billion.
The Verge AI
ChatGPT now supports generating animated images from prompts or photos, with Images 2.6 launching alongside
ChatGPT can now generate GIFs from prompts or uploaded photos, with adjustable animation speed, frame-by-frame editing, and sharing. The feature rolls out alongside OpenAI's new Images 2.6 model, which OpenAI says makes ChatGPT image generation faster; users now generate more than 4 billion images per week across ChatGPT and the API, up from the 3 billion announced in September. Images 2.6 is available to all ChatGPT and Codex users across web, desktop, and mobile, while the API side simultaneously launches GPT-Image-2.6 Flare and Sunburst.
Heat trend for this story
- Current
- 37
- Period peak
- 187 Nov 5 20:00
- 24-hour change
- -71%
Move the pointer or select the chart to inspect hourly heat. Use the left and right arrow keys with a keyboard.