BrowseComp — BrowseComp EN with context-manager tooling; distinct from ZH counterpart
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Gemini 3.1 Pro | 85.9 | 2 | 2026-05-18 | |
| 2 | Claude Opus 4.6 | Anthropic | 84 | 2 | 2026-05-18 |
| 3 | GPT-5.4 | OpenAI | 82.7 | 2 | 2026-05-18 |
| 4 | GLM-5.1 | Z.ai | 79.3 | 2 | 2026-05-18 |
| 5 | GLM-5 | Z.ai | 75.9 | 3 | 2026-05-18 |
| 6 | Kimi K2.5 | Moonshot | 74.9 | 6 | 2026-05-30 |
| 7 | Step 3.5 Flash | StepFun | 69 | 3 | 2026-05-30 |
| 8 | Claude Opus 4.5 | Anthropic | 67.8 | 1 | 2026-05-17 |
| 9 | DeepSeek-V3.2 | DeepSeek | 67.6 | 9 | 2026-06-13 |
| 10 | GLM-4.7 | Z.ai | 67.5 | 5 | 2026-05-30 |
| 11 | GPT-5.2 | OpenAI | 65.8 | 1 | 2026-05-17 |
| 12 | MiniMax M2.1 | MiniMax | 62 | 3 | 2026-05-30 |
| 13 | Kimi K2 (Reasoning) | Moonshot | 60.2 | 5 | 2026-07-09 |
| 14 | Gemini 3 Pro | 59.2 | 4 | 2026-06-13 | |
| 15 | MiMo V2 Flash | Xiaomi | 58.3 | 5 | 2026-06-13 |
| 16 | GLM-4.6 | Z.ai | 57.5 | 1 | 2026-05-18 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.