OpenAI has rolled out the Images 2.6 model to ChatGPT and Codex users to enable single-step animated image creation with speed and frame editing tools, amid the generation of more than 4 billion images weekly.
Launch of animated image creation feature with new Images 2.6 model
Playback speed control, frame editing, and character consistency preservation
Gradual rollout and release of Flare and Sunburst versions for developers
Surge in image generation to 4 billion weekly and competition with Giphy and Tenor
Analytical reading of the event’s strategic and economic dimensions
FAQs
Launch of animated image creation feature with new Images 2.6 model
The ChatGPT chatbot can now create animated GIFs directly from a user-written text prompt or an uploaded photograph. This new feature comes with OpenAI’s updated Images 2.6 model, which has begun rolling out to all ChatGPT and Codex tool users across web, desktop, and mobile applications, according to tech publication The Verge on Wednesday, November 4, 2026.
Once an animated image is generated, users can change its motion speed, edit individual frames within it, and share it directly. Prior to this update, ChatGPT was not capable of creating an original animated image in a single step, and OpenAI confirms that the new model also makes generating static images faster than before.
Playback speed control, frame editing, and character consistency preservation
OpenAI demonstrated the new feature using the prompt: “Can you make a GIF of a cat dancing?” ChatGPT returned an animation equipped with controls to change playback speed and navigate through it frame by frame, according to The Verge. Android Authority reported that the playback speed can be adjusted from half speed (0.5x) to double speed (2x), noting that ChatGPT’s previous attempts merely simulated motion superficially by placing two slightly different frames side by side.
The Decoder noted that the demonstration links minor variations of the same character in a flipbook style, confirming that the model maintains enough character feature consistency to remain recognizable across all frames—something previous models frequently failed to do. According to Crypto Briefing, a code modification integrated by OpenAI into the Codex code repository on November 3 enables API output of animated images ranging from 2 to 16 frames, with a default setting of 9 frames displayed at 200 milliseconds each.
Gradual rollout and release of Flare and Sunburst versions for developers
The feature’s rollout to users appears to be gradual; Decrypt reported that the feature was not yet available on its accounts during testing, and OpenAI has not published a separate detailed announcement about the model or specified a timeline for its arrival on all accounts.
For developers, OpenAI is launching GPT Image 2.6 Flare and GPT Image 2.6 Sunburst versions in the API. The Flare version targets faster generation, while the Sunburst version focuses on higher-quality images according to The Verge, marking the same two categories first introduced with version 2.5. Meanwhile, ShiftDelete noted that the public API pricing page was still displaying version 2.5 pricing as of Wednesday evening.
Surge in image generation to 4 billion weekly and competition with Giphy and Tenor
OpenAI announced that users are now generating more than 4 billion images weekly across the ChatGPT platform and its API, up from 3 billion images weekly announced last September according to The Verge.
Android Authority suggested that accelerating the generation process and supporting animated images will push this figure to even higher levels, placing ChatGPT in direct competition with popular animated image libraries like Giphy and Tenor in the group chat reaction image market.
Analytical reading of the event’s strategic and economic dimensions
Equipping ChatGPT with a native tool to create animated images via the new Images 2.6 model reflects OpenAI’s continuous pursuit to transform its chatbot into an integrated visual creative platform that meets the needs of fast daily communication. While previous models suffered from a lack of character feature consistency between consecutive frames, the new model has succeeded in generating coherent animation loops ranging from 2 to 16 frames from a single text prompt or uploaded image, while granting users direct tools to control playback speed and edit each frame individually.
On the commercial and competitive level, the rollout of the fast Flare and high-quality Sunburst versions for developers comes at a time when the volume of images generated via OpenAI’s ecosystem has jumped from 3 billion to more than 4 billion images weekly within just two months. Combining generation speed with the ease of producing and sharing animated images is set to drive this figure to new record highs, positioning ChatGPT in direct competition with established platforms like Giphy and Tenor in the market for reaction and expression imagery in digital conversations.
FAQs
Question: What new feature did the Images 2.6 model add to ChatGPT?
Answer: It added the ability to create animated GIF images in a single step from text prompts or uploaded images, with the option to edit frames and control playback speed from 0.5x to 2x.
Question: How many frames does the new system support when generating animated images?
Answer: Code modification in the images API supports from 2 to 16 frames for an animated image, with a default setting of 9 frames displayed for 200 milliseconds each.
Question: What are the two developer-targeted versions in the Images 2.6 model?
Answer: The company launched GPT Image 2.6 Flare dedicated to rapid generation, and GPT Image 2.6 Sunburst dedicated to producing ultra-high-quality images.
Question: How many images do ChatGPT users generate weekly?
Answer: Generation volume has exceeded 4 billion images weekly across ChatGPT and the API, compared to about 3 billion images weekly last September.