DeepSeek’s First-generation Reasoning Models

DeepSeek’s first-generation thinking designs, achieving efficiency similar to OpenAI-o1 throughout math, code, and reasoning tasks.

Models

DeepSeek-R1

Distilled models

DeepSeek team has shown that the reasoning patterns of bigger designs can be distilled into smaller sized models, resulting in better efficiency compared to the thinking patterns discovered through RL on small models.

Below are the models created by means of numerous thick designs commonly used in the research study neighborhood utilizing thinking data generated by DeepSeek-R1. The assessment results show that the distilled smaller sized dense models carry out extremely well on criteria.

DeepSeek-R1-Distill-Qwen-1.5 B

DeepSeek-R1-Distill-Qwen-7B

DeepSeek-R1-Distill-Llama-8B

DeepSeek-R1-Distill-Qwen-14B

DeepSeek-R1-Distill-Qwen-32B

DeepSeek-R1-Distill-Llama-70B

License

The design weights are certified under the MIT License. DeepSeek-R1 series support commercial usage, allow for any adjustments and derivative works, including, but not limited to, distillation for training other LLMs.