DeepSeek-R1
deepseek-thinking
Classic open reasoning model for transparent math, coding, and deliberate problem solving
ReasoningTool CallingOpen Weights
Context Window
128K
Max Output
33K
Temperature
Yes
Open Weights
Yes
Knowledge Cutoff
2024-07
Released
2025-01-20
Last Updated
2025-05-29
Modalities
Input:Text
→Output:Text
Benchmarks
| Benchmark | Score | Metric |
|---|---|---|
| Aider Polyglot | 56.9 | percent correct |
| Artificial Analysis Coding Index | 15.9 | index |
| SciCode | 35.7 | percent correct |
| Terminal-Bench Hard | 6.1 | success rate |
Available Providers (5)
| Provider | Input /1M | Output /1M | Cache Read /1M | Cache Write /1M | Reasoning | Status |
|---|---|---|---|---|---|---|
 Vercel AI Gateway | Rp 24.346 | Rp 97.382 | — | — | ✓ | — |
 AnyAPI | — | — | — | — | ✓ | — |
 GitHub Models | Gratis | Gratis | — | — | ✓ | — |
 Kilo Gateway | Rp 12.624 | Rp 45.085 | — | — | ✓ | — |
 OpenRouter | Rp 12.624 | Rp 45.085 | — | — | ✓ | — |