GPT-5.1 Codex Mini: benchmark results
OpenAI's compact Codex-specialized GPT-5.1 tier. Provider: OpenAI. Released 2025-11-13. Access: API.
Aggregate Rating 1663 ± 16, rank #286 of 1656 rated models, from 92 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SnakeBench | 36.5 | TrueSkill Rating | 98.4 |
| AA LiveCodeBench | 83.6 | Pass@1 (%) | 94.7 |
| AA AIME 2025 | 91.67 | Accuracy (%) | 93.2 |
| AA IFBench | 67.89 | Accuracy (%) | 81.4 |
| Tinybird AI SQL Benchmark - First-Attempt Success Rate | 98 | Questions answered with a valid query on the first attempt ( | 80.5 |
| Tinybird AI SQL Benchmark - Exactness | 53.65 | Result exactness vs human reference queries (0-100) | 79.3 |
| LM Market Cap LMC Score | 88.7 | LMC Score (0-100) | 78 |
| AA MMLU-Pro | 82 | Accuracy (%) | 77.9 |
| AA-Omniscience Hallucination Rate | 51.58 | Hallucination Rate (%) | 77.9 |
| AA Terminal-Bench Hard | 33.33 | Accuracy (%) | 77.7 |
| Terminal-Bench 2.0 | 61.57 | Accuracy (%) | 75 |
| AA Global-MMLU-Lite - Yoruba | 69.92 | Accuracy (%) | 74.5 |
Interactive version: theaggregate.ai/model?slug=gpt-5-1-codex-mini · How It Works · Data refreshed daily, snapshot 2026-10-20.