Verified record · from the answer index
deepseek-ai/deepseek-r1
modeltested · static-verifiedscore 85/100signed passport ✓published 182d ago · verified 2026-10-242,300,000 uses
A reasoning model trained with reinforcement learning, on par with OpenAI o1
Declared tools (1)
deepseek-ai/deepseek-r1
model:replicate:deepseek-ai/deepseek-r1