DeepSeek R1 (Jan '25): benchmark results
Provider: DeepSeek. Released 2025-01-20. Access: Open.
Aggregate Rating 1626 ± 41, rank #373 of 1656 rated models, from 68 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA MMLU-Pro | 84.45 | Accuracy (%) | 89.4 |
| AA-Omniscience Accuracy | 30.52 | Accuracy (%) | 68.6 |
| AA LiveCodeBench | 61.69 | Pass@1 (%) | 67.7 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | -4 | Omniscience Index | 66.1 |
| AA-Omniscience Index - Software Engineering (SWE) - TypeScript | -3.33 | Omniscience Index | 63.6 |
| AA AIME 2025 | 68 | Accuracy (%) | 62.4 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 32 | Omniscience Index | 62.4 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 14 | Omniscience Index | 62 |
| AA-Omniscience Index - Software Engineering (SWE) - JavaScript | -7.27 | Omniscience Index | 60.7 |
| AA-Omniscience Index - Software Engineering (SWE) - PHP | -14 | Omniscience Index | 59.4 |
| AA-Omniscience Index - Software Engineering (SWE) | -14.8 | Omniscience Index | 59.2 |
| AA-Omniscience Index - Software Engineering (SWE) - Julia | -32 | Omniscience Index | 57.3 |
Interactive version: theaggregate.ai/model?slug=deepseek-r1-jan-25 · How It Works · Data refreshed daily, snapshot 2026-10-20.