| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Opus 5 | Anthropic | 57.8 | 1 | 2026-07-24 |
| 2 | GPT-5 | OpenAI | 57.7 | 1 | 2026-07-09 |
| 3 | GPT-5.2 | OpenAI | 56.8 | 1 | 2026-07-09 |
| 4 | GPT-5.5 | OpenAI | 56.5 | 1 | 2026-07-09 |
| 5 | GPT-5.6 | OpenAI | 55.8 | 1 | 2026-07-09 |
| 6 | GPT-5.4 | OpenAI | 54 | 1 | 2026-07-09 |
| 7 | GPT-5.1 | OpenAI | 50.9 | 1 | 2026-07-09 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.