Gemini 4 Argon
Claude Opus 5.5
Muse Wonder
MiMo-V2.6-Pro
Qwen3.8-Max
Kimi K3
GLM-5.3
GPT-5.5
ERNIE 5.1
Grok 4.20 Beta
amazon-nova-experimental-chat
Dola Seed 2.0 Pro
NVIDIA-Nemotron-3-Ultra-550B-A55B
Gemma 4 31B
Step 5 Preview
DeepSeek-R1
DeepSeek-R1 is an open-weight model from DeepSeek, released May 28, 2025. It ranks #101 of 351 on the LMArena text leaderboard with a rating of 1428. It costs $0.50 per million input tokens and $2.15 per million output tokens, with a 160K-token context window.
- Developer
- DeepSeek
- Released
- May 28, 2025
- Context window
- 160K
- Price in / out, per 1M tokens
- $0.50 / $2.15
- Weights
- Open · MIT
- Inputs
- Text
- Reasoning
- Yes
- Tool use
- Yes
- Hugging Face downloads, 30 days
- 165.2K
- Hugging Face likes
- 2.5K
The rating LMArena published at each update since Jan 24, 2025. Ratings move as votes come in and new models join.
Benchmarks
| Benchmark | Score bar | Accuracy | Rank | Cost / test |
|---|---|---|---|---|
| LiveCodeBenchCoding | 70.22% | 89 / 143 | $1.35 / $5.4 | |
| MedQAMedicine | 90.80% | 43 / 92 | $1.35 / $5.4 | |
| MMLU-ProReasoning | 83.18% | 75 / 138 | $1.35 / $5.4 | |
| AIMEReasoning | 73.96% | 52 / 96 | $1.35 / $5.4 |
Vals AI, as of Oct 24, 2026.
LMArena, by category
| Leaderboard | Rating | Rank | Votes |
|---|---|---|---|
| TextLMArena | 1428 | 101 / 351 | 18.2K |
| CodingLMArena | 1427 | 141 / 346 | 2.7K |
| MathLMArena | 1400 | 161 / 339 | 861 |
| Creative writingLMArena | 1405 | 84 / 349 | 2.4K |
| Instruction followingLMArena | 1381 | 155 / 351 | 4K |
| Hard promptsLMArena | 1415 | 141 / 351 | 6.9K |
| Multi-turnLMArena | 1404 | 149 / 349 | 2.8K |
| Long queriesLMArena | 1391 | 164 / 330 | 3.1K |
| ExpertLMArena | 1394 | 169 / 305 | 887 |
| Legal and governmentIndustries | 1419 | 133 / 325 | 1.4K |
| Medicine and healthcareIndustries | 1445 | 86 / 325 | 1.2K |
| Business and financeIndustries | 1401 | 137 / 344 | 2.9K |
| ChineseLanguages | 1442 | 139 / 328 | 1.2K |
| SpanishLanguages | 1412 | 131 / 249 | 261 |
| GermanLanguages | 1404 | 122 / 261 | 597 |
| FrenchLanguages | 1417 | 138 / 252 | 237 |
| JapaneseLanguages | 1391 | 88 / 236 | 621 |
LMArena, CC BY 4.0, as of Oct 22, 2026. Where a model is rated at several effort levels, its best is shown (DeepSeek-R1 (0528), DeepSeek-R1).
Compare DeepSeek-R1
DeepSeek
A research lab known for strong open-weight reasoning and coding models trained at low cost. Founded 2023, based in Hangzhou. deepseek.com
Questions
Who makes DeepSeek-R1?
DeepSeek-R1 is made by DeepSeek, founded 2023 and based in Hangzhou.
How much does DeepSeek-R1 cost?
$0.50 per million input tokens and $2.15 per million output tokens, at its provider’s list price.
What is DeepSeek-R1’s context window?
160K tokens (163,840).
How does DeepSeek-R1 rank?
#101 of 351 on LMArena’s text leaderboard, rated 1428 (95% interval 1422–1433) from 18,170 votes, as of Oct 22, 2026.
Is DeepSeek-R1 open source?
DeepSeek-R1 has open weights under the MIT license, downloadable from Hugging Face (deepseek-ai/DeepSeek-R1-0528).