| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | GPT-5.6 Sol | OpenAI | 52.7 | 2 | 2026-08-23 |
| 2 | Qwen3.8 Max Preview | Alibaba | 52.4 | 1 | 2026-08-23 |
| 3 | GPT-5.6 Terra | OpenAI | 50.4 | 1 | 2026-08-23 |
| 4 | GPT-5.6 Luna | OpenAI | 50.3 | 1 | 2026-08-23 |
| 5 | Qwen3.8 27B | Alibaba | 42.9 | 2 | 2026-08-24 |
| 6 | Qwen3.7 Plus Preview | Alibaba | 33.6 | 1 | 2026-08-24 |
| 7 | GLM-5.3 | Z.ai | 28.5 | 1 | 2026-08-23 |
| 8 | Kimi K3 | Moonshot | 28.3 | 2 | 2026-08-13 |
| 9 | DeepSeek-V4-Flash-Vision-Exp | DeepSeek | 27.3 | 1 | 2026-08-23 |
| 10 | Qwen3.6 27B | Alibaba | 27.3 | 1 | 2026-08-24 |
| 11 | Claude Opus 4.8 | Anthropic | 27 | 3 | 2026-08-13 |
| 12 | GPT-5.5 | OpenAI | 26.6 | 1 | 2026-07-27 |
| 13 | Gemini 3.7 Flash | 26.3 | 2 | 2026-08-24 | |
| 14 | Claude Fable 5 | Anthropic | 25.7 | 1 | 2026-07-27 |
| 15 | DeepSeek-V4-Flash | DeepSeek | 25.2 | 6 | 2026-08-23 |
| 16 | GLM-5.2 Full Open Source | Z.ai | 23.8 | 3 | 2026-08-13 |
| 17 | DeepSeek-V4-Pro | DeepSeek | 16.5 | 2 | 2026-08-13 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.