GPT-5.1 (High): benchmark results
GPT-5.1 evaluated at the high reasoning-effort setting. Provider: OpenAI. Released 2025-11-12. Access: API.
Unified ELO 1668 ± 1, rank #185 of 1761 rated models, from 165 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Global-MMLU-Lite - French | 93.5 | Accuracy (%) | 99.1 |
| Vals AI MedQA | 96.38 | Accuracy (%) | 98.9 |
| Vals AI CaseLaw v2 | 73.42 | Accuracy (%) | 98.3 |
| AA LiveCodeBench | 86.77 | Pass@1 (%) | 97.5 |
| CLBench | 23.7 | Solving Rate (%) | 97.1 |
| AA Omniscience - Software Engineering (SWE) - R | 42 | Accuracy (%) | 96.8 |
| AA MMLU-Pro | 87.04 | Accuracy (%) | 96.7 |
| AA Global-MMLU-Lite - English | 94.08 | Accuracy (%) | 96.4 |
| LLM2014 Logic 2025-11 | 79.53 | Median Score | 96.2 |
| AA AIME 2025 | 94 | Accuracy (%) | 96.1 |
| AA Omniscience - Software Engineering (SWE) - JavaScript | 55.45 | Accuracy (%) | 96.1 |
| AA Omniscience - Software Engineering (SWE) - Swift | 64 | Accuracy (%) | 96.1 |
Interactive version: theaggregate.ai/model?slug=gpt-5-1-high · How It Works · Data refreshed daily, snapshot 2026-09-05.