AIME (year unspecified) — Anthropic Claude 4 announcement gives no year; cannot safely assign to 2024/2025/2026 vintage
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.7 | Anthropic | 91 | 1 | 2026-07-05 |
| 2 | Gemini 2.5 Flash | 56.3 | 1 | 2026-07-05 | |
| 3 | MiMo V2.5 Pro | Xiaomi | 37.3 | 1 | 2026-08-23 |
| 4 | Claude 4 Opus | Anthropic | 33.9 | 1 | 2026-06-13 |
| 5 | Claude Sonnet 4 | Anthropic | 33.1 | 1 | 2026-06-13 |
| 6 | GPT OSS 120B | OpenAI | 8.6 | 1 | 2026-07-05 |
| 7 | GPT-5 | OpenAI | 7.6 | 1 | 2026-07-05 |
| 8 | Qwen2.5 Max | Alibaba | 2.3 | 1 | 2026-07-05 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.