DeepSeek-R1
DeepSeek- Arena score
- 1373
- Among open models
- #58of 148
- Among all on Arena
- #186
- Votes
- 18,524
Measured on DeepSeek-R1, the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.
Score by category
CategoryScoreOpen rankArena rankVotes
- Chinese1400#58#1821.2K
- Math1393#55#1751.6K
- Multi-turn1390#51#1672.4K
- Applied math1390#55#1791.5K
- English1385#60#18510.7K
- German1384#42#144436
- Spanish1379#54#170114
- Science1375#63#1943.3K
- Software & IT1375#67#2064K
- Hard prompts, English1375#67#2052.7K
- Coding1371#69#2082.3K
- Legal & government1371#59#1921.1K
- Medicine & health1371#60#189830
- French1366#57#181188
- Writing & literature1364#51#1745.4K
- Hard prompts1361#64#2054.1K
- Instructions1357#54#1836.4K
- Non-English1357#57#1897.8K
- Long prompts1355#60#1982.3K
- Russian1354#58#1911.9K
- Creative writing1354#44#1643.3K
- Business & finance1348#64#1991.8K
- Media & sports1347#53#1773.6K
- Excluding ties1346#56#18312.5K
- Expert1337#73#216848
- Korean1330#44#156215
- Japanese1324#41#151468
| Model | Params | Score |
|---|---|---|
| 684B | 1373 | |
| 685B | 1428 |
About as strong, smaller
Compare
gemma-4-26B-A4B-itGoogle143425.8B

gemma-3-27b-itGoogle135827.4B

Qwen3.8-27BQwen143727.8B

Qwen3.5-27BQwen140927.8B
Discussion
Built on it
Ways to run it
- GGUF · llama.cpp, Ollama, LM Studiounsloth/DeepSeek-R1-GGUF26 builds
- MLX · Apple silicon: mlx-lm, LM Studiomlx-community/DeepSeek-R1-4bit7 builds
- AWQ · vLLM, SGLangQuixiAI/DeepSeek-R1-AWQ4 builds
- GPTQ · vLLM, SGLangOPEA/DeepSeek-R1-int4-gptq-sym-inc3 builds
Most downloaded
BenchClub runs
Not runnable yet
Runs open once a DeepSeek-R1 build is registered.
More from DeepSeek
See more
DeepSeek-V3.29.6M685B

DeepSeek-V4-Flash-07312M304B

DeepSeek-V4.1-Flash1.6M763B

DeepSeek-V4-Flash1.2M291B
Other >300B models
See more
GLM-5.3-FlashZ.ai6.4M321B
InklingThinking Machines1.5M952B
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16NVIDIA1M561B

Kimi-K3Moonshot AI768.9K2.8T