OpenAI发布Images 2.6：一张照片变GIF，生图提速50%
BlockBeats称，OpenAI发布Images 2.6，支持将照片生成GIF并提升图片生成速度；Flare和Sunburst两款模型已上线API。
Loading…
bibibi.ccbibibi.ccLoading…
Connect scattered updates into a story you can understand.
BlockBeats称，OpenAI发布Images 2.6，支持将照片生成GIF并提升图片生成速度；Flare和Sunburst两款模型已上线API。
Beating AI said that developers found OpenAI API's list had displayed three Aurora codenames; an evaluation account showed image samples from the red and green versions, while their specific purpose has not yet been confirmed by OpenAI.
Black Forest Labs Open-source FLUX 3 Image Klein, supporting 4K generation, image editing, and reference-image input; official evaluations lead, but community and third-party test results are mixed.
Beating AI reports that Alibaba Tongyi Qianwen has released Qwen-Image-2.1-Turbo and opened its model weights, reducing the official example's denoising steps from 40 steps to 8 steps. The weights are licensed for non-commercial research and evaluation; commercial development requires separate authorization.
The Alibaba Qwen team released Qwen-Image-2.1-Turbo, supporting 2K image generation and continuous editing through natural language; the model weights and related API have been made available.
PANews, citing the Google DeepMind model card, reports that Google has released the Nano Banana 2.1 image generation and editing model, now available in the Gemini app and on API.
Tencent Hunyuan has launched Hy Image3.5 preview, increasing the reference-image limit to 20 and supporting multi-round editing and direct 4K output. The model is now available on Miora and is free for a limited time from September 22 to October 7.