As open-source models quickly evolve, we are focused on providing the most cutting-edge options to our users.
The DeepSeek V4 Pro serverless Model APIs is scheduled to be retired at 5pm PT on September 23, at which time this model will be disabled and any requests made to it will return HTTP error code 400.
As a replacement, we suggest using DeepSeek V4.1 Flash, which delivers stronger performance at a lower price, or GLM 5.3, which delivers frontier-level performance. Before the cutoff, update your integration to use lightning-ai/deepseek-v4.1-flash or lightning-ai/glm-5.3. If you still need access to DeepSeek V4 Pro, please reach out to discuss setting up a dedicated deployment.