GPT-5.6 Sol (Non-reasoning): benchmark results
GPT-5.6 Sol evaluated with reasoning disabled. Provider: OpenAI. Released 2026-07-09. Access: API.
Unified ELO 1653 ± 1, rank #235 of 1761 rated models, from 24 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI - Natural Intelligence | 70.1 | NatInt Score | 97.6 |
| AA Omniscience - Health | 45.4 | Accuracy (%) | 93.3 |
| AA Omniscience - Software Engineering (SWE) | 73 | Accuracy (%) | 93.2 |
| UGI - Writing | 59.94 | Writing Score | 92.7 |
| AA Omniscience - Business | 40.9 | Accuracy (%) | 91.4 |
| AA-Omniscience Accuracy | 48.7 | Accuracy (%) | 90.8 |
| AA Omniscience - Law | 43.4 | Accuracy (%) | 90.6 |
| AA Omniscience - Humanities & Social Sciences | 44.5 | Accuracy (%) | 90.3 |
| Wolfram LLM Benchmarking Project | 61.4 | Correct Functionality (%) | 89.7 |
| AA Omniscience - Science, Engineering & Mathematics | 45 | Accuracy (%) | 86.7 |
| Artificial Analysis Intelligence Index | 32.86 | Intelligence Index | 84.9 |
| AA Omniscience | 1.1 | Score | 83 |
Interactive version: theaggregate.ai/model?slug=gpt-5-6-sol-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.