GPT-5.6 Terra (xHigh): benchmark results
GPT-5.6 Terra evaluated at the xhigh reasoning-effort setting. Provider: OpenAI. Released 2026-07-09. Access: API.
Aggregate Rating 1826 ± 10, rank #94 of 2168 rated models, from 187 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Terminal-Bench Hard | 62.88 | Accuracy (%) | 99.5 |
| Vals AI TaxEval v2 | 76.17 | Accuracy (%) | 96.2 |
| AA CritPt | 27.14 | Accuracy (%) | 94.2 |
| ProofBench | 71 | Score (imported) | 93.9 |
| Chatbot Arena (Text - Mathematical) | 1491 | Arena Score | 92.3 |
| Chatbot Arena (Text - Japanese) | 1475 | Arena Score | 92.1 |
| AA-Briefcase - Presentation Elo | 1507.32 | Elo | 91.8 |
| LLM Chess (Saplin) | 1293.01 | Native anchored-opponent Elo (signed points) | 91.5 |
| Chatbot Arena (Text - Math) | 1482 | Arena Score | 91.2 |
| Chatbot Arena (Text - German) | 1478 | Arena Score | 91 |
| AA GPQA Diamond | 90.81 | Accuracy (%) | 90.9 |
| Chatbot Arena (Text - Expert) | 1504 | Arena Score | 90.9 |
Interactive version: theaggregate.ai/model?slug=gpt-5-6-terra-xhigh · How It Works · Data refreshed daily, snapshot 2026-10-20.