DeepSeek Launches Multimodal Model V4-Flash-Vision-Exp, Opens API Services

Odaily News: According to the DeepSeek official WeChat account, DeepSeek has launched the multimodal visual understanding model V4-Flash-Vision-Exp and opened up API services.

Building on V4-Flash, this model adds image understanding and visual generation capabilities, supports multimodal input and output including text and images, and provides developers with interfaces for calling capabilities such as inference and generation.