GPT-5.1 Codex Max: benchmark results
OpenAI's higher-compute GPT-5.1 Codex variant for coding and software-agent tasks. Provider: OpenAI. Released 2025-11-19. Access: API.
Aggregate Rating 1763 ± 15, rank #100 of 1656 rated models, from 64 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MERA Code | 0.54 | Total score | 100 |
| MERA Code - stRuCom | 38.3 | chrF (%) | 100 |
| SnakeBench | 38.2 | TrueSkill Rating | 100 |
| MERA Code - CodeCorrectness | 91.33 | Accuracy (%) | 98.5 |
| MERA Code - JavaTestGen | 55.95 | pass@1 (%) | 98.5 |
| MERA Code - RealCode | 48.5 | pass@1 (%) | 98.5 |
| MERA Code - UnitTests | 32.31 | CodeBLEU (%) | 96.9 |
| MonitoringBench - Human Strategies - Malware Download | 100 | Catch rate (%) at a 1% false-positive-rate threshold calibra | 95.8 |
| MERA Code - CodeLinterEval | 95.55 | pass@1 (%) | 95.4 |
| MERA Code - ruCodeReviewer | 12.34 | Top-1 Accuracy (%) | 93.8 |
| MonitoringBench - Human Strategies - Password Leak | 98.4 | Catch rate (%) at a 1% false-positive-rate threshold calibra | 91.7 |
| MonitoringBench - Human Strategies - Sudo Escalation | 84 | Catch rate (%) at a 1% false-positive-rate threshold calibra | 91.7 |
Interactive version: theaggregate.ai/model?slug=gpt-5-1-codex-max · How It Works · Data refreshed daily, snapshot 2026-10-20.