deepseek-ai/deepseek-r1

A reasoning model trained with reinforcement learning, on par with OpenAI o1

2.1M runs
License
GitHub
Weights
Playground API Examples README