DeepSeekDeepSeek R1Open SourceArena #42May 28, 2025
DeepSeek: R1 0528 (free)
May 28th update to the original DeepSeek R1 Performance on par with OpenAI o1, but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass. Fully open-source model.
Context Window
164K
tokens
Max Output
164K
tokens
Released
May 28, 2025
Arena Rank
#42
of 305 models
Output Speed
135
tokens/sec
Time to First Token
4.8s
TTFT
Capabilities
👁Vision
🧠Reasoning
🔧Tool Calling
⚡Prompt Caching
🖥Computer Use
🎨Image Generation
Supported Parameters
Frequency Penalty
Reduce repetition
Include Reasoning
Show reasoning tokens
Max Tokens
Output length limit
min_p
Presence Penalty
Encourage new topics
Reasoning
Extended thinking
Repetition Penalty
Penalize repeated tokens
Stop Sequences
Custom stop tokens
Temperature
Controls randomness
Top K
Top-K sampling
Top P
Nucleus sampling
Pricing Comparison
| Router | Input / 1M | Output / 1M | Cached Input / 1M |
|---|---|---|---|
| OpenRouter | Free | Free | — |
| Martian | $0.40 | $1.75 | $0.20 |
| DeepInfra | $1.00 | $3.00 | — |
Benchmarks
Artificial Analysis
Intelligence IndexArtificial Analysis
70/100Coding IndexArtificial Analysis
62/100Math IndexArtificial Analysis
94/100MMLU-PROArtificial Analysis
0.798/1GPQA DiamondArtificial Analysis
0.718/1MATH-500Artificial Analysis
0.972/1AIME 2024Artificial Analysis
0.897/1Humanity's Last ExamArtificial Analysis
0.088/1LiveCodeBenchArtificial Analysis
0.655/1SciCodeArtificial Analysis
0.408/1Model IDs
OpenRouter
deepseek/deepseek-r1-0528:freeHugging Facedeepseek-ai/DeepSeek-R1-0528 ↗
Tags
reasoning