GPT-6.1 Sol Pro
OpenAI · 2026-09-29
GPT-6.1 Sol Pro is the higher-compute Pro execution mode of GPT-6.1 Sol, selected through the gpt-6.1-sol API model with reasoning.mode=pro. Atlas keeps explicitly Pro-labeled benchmark results separate from standard execution, following its vendor-named Pro identity policy.
Benchmark scores
| Benchmark | Score |
|---|---|
| BuseyBench SVG | 7.1 |
| MineBench — Bradley-Terry Rating | 2134 |