DeepSeek-V2.5-1210
DeepSeek- Arena score
- 1294
- Among open models
- #85of 148
- Among all on Arena
- #257
- Votes
- 6,795
Measured on DeepSeek-V2.5 (1210), the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.
Score by category
CategoryScoreOpen rankArena rankVotes
- Chinese1318#79#241446
- Software & IT1313#85#2541.7K
- English1310#90#2613.8K
- Coding1309#85#2551.1K
- Long prompts1301#82#2501K
- Hard prompts, English1299#87#2571.1K
- Multi-turn1297#83#2531.2K
- Medicine & health1292#87#253291
- Russian1289#80#2431K
- Hard prompts1289#86#2571.8K
- Math1288#83#2471K
- Science1288#92#2701.1K
- Creative writing1285#76#2411.1K
- Writing & literature1283#82#2551.9K
- Instructions1279#83#2503K
- Legal & government1279#97#275375
- Applied math1278#87#262893
- Media & sports1275#79#2431.2K
- Non-English1273#85#2573K
- Expert1266#85#250441
- Business & finance1259#91#268715
- German1258#73#230140
- Excluding ties1229#87#2594.2K
- Japanese1228#58#202161
| Model | Params | Score |
|---|---|---|
| 236B | 1271 | |
| 236B | 1294 |
About as strong, smaller
Compare
granite-4.2-3bIBM Granite12923.7B
gemma-3-4b-itGoogle12914.3B

gemma-3n-E4B-itGoogle13067.8B

Molmo2-8BAi212878.7B
Built on it
Ways to run it
- GGUF · llama.cpp, Ollama, LM Studiobartowski/DeepSeek-V2.5-1210-GGUF5 builds
- MLX · Apple silicon: mlx-lm, LM Studiomlx-community/DeepSeek-V2.5-1210-4bit5 builds
- AWQ · vLLM, SGLangadamo1139/DeepSeek-V2.5-1210-AWQ1 build
Most downloaded
- DeepSeek-V2.5-1210-GGUFbartowski · Quantization1.8K downloads
- DeepSeek-V2.5-1210-4bitmlx-community · Quantization283 downloads
- DeepSeek-V2.5-1210-GGUFlmstudio-community · Quantization232 downloads
- deepseek-ai.DeepSeek-V2.5-1210-GGUFDevQuasar-5 · Quantization102 downloads
- DeepSeek-V2.5-1210-3bitmlx-community · Quantization55 downloads
BenchClub runs
Not runnable yet
Runs open once a DeepSeek-V2.5-1210 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 80–300B models
See more
MiniMax-M2.5MiniMax1M229B
Inkling-SmallThinking Machines1M266B
Hy3Tencent987.2K299B

NVIDIA-Nemotron-3-Super-120B-A12B-BF16NVIDIA874.7K124B