DeepSeek-V4-Pro Officially Launches With 1.6T Params And 1M Context

DeepSeek releases DeepSeek-V4-Pro-0813 officially: a 1.6T-parameter MoE activating 49B per token with 1M-token context, aimed at long-horizon agent work, alongside an open-source eval harness. CoreWeave also launched serverless inference for the model.

2026-09-02 ~ 2026-09-03 · 2 related posts

• DeepSeek-V4-Pro launches on CoreWeave with 1.6T params, 1M context — wandb · 2026-09-02
• DeepSeek-V4-Pro ships with 1.6T-param MoE; open-source eval harness steals the show — DeepLearningAI · 2026-09-03