Natural Questions — Tencent Hunyuan source, scope unstated; kept apart from closedbook despite close band.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Hunyuan Large | Tencent | 52.8 | 1 | 2026-05-31 |
| 2 | Llama 3.1 Instruct 405B | Meta | 41.5 | 1 | 2026-05-01 |
| 3 | DeepSeek-V3 | DeepSeek | 40 | 1 | 2026-05-01 |
| 4 | DeepSeek-V2 | DeepSeek | 38.7 | 2 | 2026-05-31 |
| 5 | Gemma 2 27B IT | 34.5 | 1 | 2026-08-23 | |
| 6 | Qwen2.5 Instruct 72B | Alibaba | 33.2 | 1 | 2026-05-01 |
| 7 | Gemma 2 Instruct (9B) | 29.2 | 1 | 2026-08-23 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.