OpenAI has taken its next step in image generation. Images 2.6, now accessible through ChatGPT, Codex and the API, brings significant upgrades in both speed and motion.
According to the company, more than 4 billion images are generated worldwide every week, averaging over half a billion per day. Images 2.6 aims to maintain OpenAI's lead in this massive volume by adding new capabilities.
Generation time cut in half
The new model creates images 50% faster than previous versions. For developers using the API, OpenAI is offering two different variants that can be chosen based on need:
- GPT-Image-2.6 Flare: Can generate an image in just 6 seconds in low quality mode. It offers strong layout quality for posters, product images and professional edits.
- GPT-Image-2.6 Sunburst: The choice for medium and high quality tiers. It delivers high detail while maintaining visual consistency across edits.
GIF support brings ideas to life
One of the most eye-catching features of Images 2.6 is the ability to turn static images into GIFs. Marketers and content creators can use this to animate brand mascots, create animated reaction images for social media, or produce animated drafts before a video shoot.
Additionally, without requiring complex drawing skills, even a simple sketch can be turned into a moving, lively character for games and videos.
Consistency in characters and products stands out
The model draws attention with its ability to maintain consistency when editing characters, products and brand details. This feature provides a significant advantage, especially for teams that want to preserve their brand identity.
Who can use it?
Images 2.6 has been rolled out globally across all subscription tiers of ChatGPT and Codex on web, desktop and mobile platforms.
Developers, meanwhile, can access the GPT-Image-2.6 Sunburst and Flare models right away through the API.
Comments (0)
No comments yet. Be the first!