DeepSeek-R1
DeepSeek-R1 is an open-weights AI model from DeepSeek with an index score of 130 (GPT-4 = 100), rank 131 of 217.
130114–146Index score#131 of 217 models
–Price per million tokensno public API price found
–Context windowtokens it reads at once
20 Jan 2025ReleasedReplaced by DeepSeek-R1-0528
Where it sitson the frontier
o1 reached this score 2 months earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingevidence
Agentsevidence
Professional workno evidence yet
Novel problemsno evidence yet
Long tasksevidence
Third-party results only, at the model’s best standard setting. Hover a bar for who ran it.
What others sayrank on 8 leaderboards
The Model Indexthird-party tests, our method#131of 217 · 130
LMArena · Textoverall chat quality, judged by people#107of 362 · 1422LMArena · Creative writingcreative writing#109of 360 · 1389LMArena · Codingcoding questions in chat#113of 357 · 1464LMArena · Mathsmaths questions in chat#119of 346 · 1412LMArena · Instruction followingfollowing detailed instructions#127of 362 · 1397LMArena · Long promptslong, complex requests#131of 340 · 1408Epoch Capabilities Indexoverall capability from many benchmarks#116of 231 · 140.8Best attop three among current models
Replaced by DeepSeek-R1-0528; it no longer leads any category among current models.
Where to use itprices checked 18 Nov 2026
Host prices from OpenRouter; going direct can differ.
Facts and sources
Released20 Jan 2025
Announcementapi-docs.deepseek.com ↗
Release entryDeepSeek-R1
Scored asDeepSeek-R1
Output speed–
Hugging Facenot linked