50% Faster Image Generation: OpenAI says Images 2.6 generates images faster while improving consistency across multiple editing rounds. New API Models and GIF Support: Flare prioritizes speed
OpenAI launches Images 2.6 with 50% faster generation, GIF support, improved editing consistency, and new API features including SVG export.
OpenAI announced the release of its latest image generation model, Images 2.6, on November 4, bringing the upgraded model to ChatGPT, Codex, and its API. According to OpenAI, the new model generates images 50% faster than its predecessor and delivers improved consistency across multiple editing rounds.
The upgrade is designed to make image creation and refinement more reliable, particularly when users need to adjust compositions, backgrounds, or visual elements while preserving the appearance of characters, product designs, and brand details.
OpenAI has introduced two API models, GPT-Image-2.6 Flare and Sunburst, each designed for different image generation needs.
Flare prioritizes speed and rapid iteration. Under low-quality settings, it can generate an image in as little as six seconds. OpenAI also highlighted improvements in layout composition, making the model more suitable for marketing materials such as posters and product images.
Sunburst, meanwhile, is designed for more detailed image generation. According to OpenAI, it delivers greater visual detail under medium- and high-quality settings while maintaining stronger consistency across multiple editing rounds.
🔍 For example, businesses can use Sunburst to create multiple advertisements featuring the same product in different environments without significantly changing the product’s original appearance.
Images 2.6 introduces GIF generation, allowing users to turn static visual concepts into short animations. Potential applications include animated brand mascots, looping product showcases, and social media memes.
OpenAI demonstrated how multiple sequential frames can be combined into looping animations, as well as how simple sketches can be developed into animated characters.
For developers, OpenAI has also brought ChatGPT’s image annotation capabilities to its API. This allows developers to mark specific areas of an image for targeted modifications, providing greater control over the editing process.
Another addition is SVG export support, giving developers more flexibility when working with scalable vector graphics.
The updated API also introduces a new complexity parameter, which controls how much planning the model applies when handling composition requirements and prompt constraints. Lower settings are suitable for straightforward edits, while medium and high settings can be used for more complicated scenes involving multiple elements or detailed layout requirements.
Image quality settings remain separate, allowing developers to independently control visual detail and fidelity.
OpenAI said Images 2.6 is rolling out globally across ChatGPT and Codex plans, covering desktop, mobile, and web platforms.
Both Flare and Sunburst are also available through the API, with pricing based on input and output tokens.
Under OpenAI’s standard pricing, image input costs $8 per million tokens, while image output costs $30 per million tokens for both models.
With faster generation, improved editing consistency, GIF creation, and expanded API capabilities, Images 2.6 represents a broader move toward making AI image generation more practical for professional creative workflows, marketing production, and developer applications.
Looking for the latest scoop and cool insights from CoinRank? Hit up our Twitter and stay in the loop with all our fresh stories!
〈OpenAI Images 2.6 Brings Faster Generation and GIFs〉這篇文章最早發佈於《CoinRank》。
