OpenRouter 上线首个图像模型 Gemini 2.5 Flash Image Preview
The First-Ever Image Model Is Up on OpenRouter
OpenRouter 上线其平台首个图像模型 Gemini 2.5 Flash Image Preview，定位为结合 LLM 智能与扩散模型视觉质量。该模型支持图像生成与编辑、跨场景角色一致性、单次调用生成多张图，上下文长度 32k，价格为每百万输入 token $0.30。
OpenRouter 官方宣布首个图像模型上线，给出了价格、上下文长度和典型工作流，读者可据此评估是否接入现有流程。
🍌 Gemini 2.5 Image Preview is Live
We’ve just launched the first-ever image model on OpenRouter: Gemini 2.5 Flash Image Preview.
This model combines the intelligence of LLMs with the visual quality of diffusion models — unlocking workflows like:
- Generating pages of a website
- Making comic panels with consistent characters
- Creating logo iterations in bulk
Features
- State-of-the-art image generation and editing
- Incredible character consistency across scenes
- Can generate multiple images in a single call
- 32k context length
Use it for just $0.30 per million input tokens → Try it here
We’re proud to be a launch partner with Google DeepMind. Learn more in the Gemini 2.5 Flash Image blog post from Google DeepMind.
Follow us on X @OpenRouterAI for updates and upcoming model features.
来源：OpenRouter：Announcements（RSS） · openrouter.ai