DeepSeek V3.1 Terminus (Reasoning) — benchmark results
DeepSeek V3.1 Terminus evaluated with reasoning enabled. Provider: DeepSeek. Released 2025-09-22. Access: Open.
Unified ELO 1722 ± 21, rank #237 of 1776 rated models, from 58 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI Leaderboard | 51.26 | UGI Score | 91.9 |
| AA LiveCodeBench | 79.79 | Pass@1 (%) | 91.8 |
| AA MMLU-Pro | 85.11 | Accuracy (%) | 91.6 |
| UGI - Writing | 54 | Writing Score | 91 |
| UGI - Natural Intelligence | 49.47 | NatInt Score | 90.7 |
| AA AIME 2025 | 89.67 | Accuracy (%) | 89.8 |
| AA Long Context Reasoning | 65 | Accuracy (%) | 81.5 |
| AA Omniscience - Humanities & Social Sciences | 29.6 | Accuracy (%) | 78.7 |
| AA Omniscience - Health | 28.7 | Accuracy (%) | 77.9 |
| AA CritPt | 1.71 | Accuracy (%) | 76.9 |
| AA Omniscience - Science, Engineering & Mathematics | 34.9 | Accuracy (%) | 76 |
| AA Omniscience - Business | 24.2 | Accuracy (%) | 75.9 |
Interactive version: theaggregate.ai/model?slug=deepseek-v3-1-terminus-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.