GPT-6.1 Sol (Max): benchmark results
Provider: OpenAI. Released 2026-09-29. Access: API.
Aggregate Rating 1962 ± 6, rank #15 of 2245 rated models, from 525 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Index - Software Engineering (SWE) - Kotlin | 85.71 | Omniscience Index | 100 |
| Chatbot Arena (Vision - Homework) | 1355 | Arena Score | 100 |
| CorporateBench | 82.21 | Mean score over 3 runs of 128 tasks (%) | 100 |
| Elorian Glimpse Card - Video-MME v2 (with subtitles) | 65.27 | Grouped rubric score | 100 |
| Epoch AI - Frontiermath Erdos | 4.41 | Score | 100 |
| MathArena - ARXIV August | 96.49 | Accuracy (%) | 100 |
| MathArena - ARXIV_FALSE September | 82.7 | Accuracy (%) | 100 |
| MultiLangBench - C | 88.4 | Test Pass Rate (%) | 100 |
| MultiLangBench - COBOL | 72.6 | Test Pass Rate (%) | 100 |
| MultiLangBench - Fortran | 84.1 | Test Pass Rate (%) | 100 |
| MultiLangBench - Go | 89.9 | Test Pass Rate (%) | 100 |
| MultiLangBench - OCaml | 87.5 | Test Pass Rate (%) | 100 |
Interactive version: theaggregate.ai/model?slug=gpt-6-1-sol-max · How It Works · Data refreshed daily, snapshot 2026-11-16.