DeepSeek V3.1 Terminus (Reasoning): benchmark results
DeepSeek V3.1 Terminus evaluated with reasoning enabled. Provider: DeepSeek. Released 2025-09-22. Access: Open.
Aggregate Rating 1621 ± 14, rank #642 of 2168 rated models, from 104 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA LiveCodeBench | 79.79 | Pass@1 (%) | 91.8 |
| AA MMLU-Pro | 85.11 | Accuracy (%) | 91.7 |
| UGI Leaderboard | 51.26 | UGI Score | 91.1 |
| AA AIME 2025 | 89.67 | Accuracy (%) | 89.9 |
| UGI - Writing | 54 | Writing Score | 88.8 |
| UGI - Natural Intelligence | 49.47 | NatInt Score | 88.4 |
| AA Omniscience - Software Engineering (SWE) - Dart | 28 | Accuracy (%) | 82.1 |
| AA Omniscience - Software Engineering (SWE) - C | 50 | Accuracy (%) | 81.2 |
| AA Omniscience - Software Engineering (SWE) - HTML | 44 | Accuracy (%) | 80.7 |
| AA Omniscience - Software Engineering (SWE) - Julia | 24 | Accuracy (%) | 80.2 |
| AA Omniscience - Software Engineering (SWE) - Go | 26 | Accuracy (%) | 78.2 |
| AA Omniscience - Software Engineering (SWE) - JavaScript | 38.18 | Accuracy (%) | 77.5 |
Interactive version: theaggregate.ai/model?slug=deepseek-v3-1-terminus-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.