DeepSeek V4 Flash Adds Vision: Multimodal Model Now Live on SiliconFlow and OpenRouter
Key Info
DeepSeek has released DeepSeek-V4-Flash-Vision-Exp, an experimental vision-enabled version of V4 Flash, now available on SiliconFlow and OpenRouter with 1M context and an MIT license.
Highlights
- Adds visual understanding to the V4 Flash model family for multimodal tasks and agent workflows.
- Maintains strong text capabilities while delivering stronger multimodal agent performance.
- Available immediately via SiliconFlow API and OpenRouter; MIT-licensed with 1M-token context.