GPT-5.6 Sol: benchmark results
OpenAI's flagship GPT-5.6 tier for demanding reasoning, coding, and agentic tasks. Provider: OpenAI. Released 2026-07-09. Access: API.
Aggregate Rating 1889 ± 3, rank #19 of 1656 rated models, from 1958 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Terminal-Bench Hard | 65.91 | Accuracy (%) | 100 |
| AI Consilium — Conclusion free-form | 2.71 | Mean clinical quality (0–3) | 100 |
| AI Consilium — Conclusion structured | 2.77 | Mean clinical quality (0–3) | 100 |
| AI Consilium — Role clinical opinion structured | 2.49 | Mean clinical quality (0–3) | 100 |
| AI Consilium — Role reasoning structured | 2.89 | Mean clinical quality (0–3) | 100 |
| AI for Education Visual Maths | 91.14 | Accuracy (%) | 100 |
| AI for Education Visual Maths - Number and Operations | 89.19 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - odd one out | 88.3 | Accuracy (%) | 100 |
| AeroCopilotBench - Tier-2 Safety-Gated Outcome | 91.1 | Goal attainment (%; zero on any safety violation) | 100 |
| AeroCopilotBench - Tier-2 Safety-Gated Success Rate | 72.6 | Success rate (%; goals met with no safety violation) | 100 |
| AgentIdeaBench - Static | 6.9 | Weighted Hypothesis Score (1-10) | 100 |
| AgentIdeaBench - Static - Clarity | 7.79 | Clarity Critic Score (1-10) | 100 |
Interactive version: theaggregate.ai/model?slug=gpt-5-6-sol · How It Works · Data refreshed daily, snapshot 2026-10-20.