Back to Models
DeepSeek AI
Deepseek R1
Open source3 variants
DeepSeek's flagship 671B reasoning model with chain-of-thought, matching o1 performance on math and coding benchmarks.
164K tokensFree / Open weightsMoEMIT
DeepSeek-R1
76.4%SU AI Zero (Anthropic+Google+OpenAI, Suzhou AI Lab)
reasoning671B
DeepSeek's May 2025 R1 update — stronger reasoning across math, coding, and science over the original R1 release.
164K tokensFree / Open weightsMoE
reasoning671B
DeepSeek's pure-RL reasoning model trained without SFT — demonstrates emergent chain-of-thought through reinforcement learning alone.
164K tokensFree / Open weights