Toolathlon — Distinct named tool-use benchmark; sampling:pass@3 variant
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Muse Spark 1.1 | Meta | 75.6 | 1 | 2026-08-23 |
| 2 | Kimi K3 | Moonshot | 73.2 | 1 | 2026-08-23 |
| 3 | GLM-5.3 | Z.ai | 73 | 2 | 2026-08-23 |
| 4 | Qwen3.8 Max Preview | Alibaba | 72.5 | 1 | 2026-08-23 |
| 5 | DeepSeek-V4-Flash | DeepSeek | 70.3 | 5 | 2026-08-23 |
| 6 | Claude Fable 5 | Anthropic | 61.7 | 1 | 2026-06-30 |
| 7 | Claude Mythos 5 | Anthropic | 61.1 | 1 | 2026-06-30 |
| 8 | Claude Opus 4.8 | Anthropic | 59.9 | 2 | 2026-08-23 |
| 9 | Claude Opus 4.7 | Anthropic | 59.3 | 1 | 2026-06-30 |
| 10 | GPT-5.6 Sol | OpenAI | 58 | 1 | 2026-08-23 |
| 11 | Claude Opus 4.6 | Anthropic | 56.8 | 6 | 2026-07-04 |
| 12 | Gemini 3.5 Flash | 56.5 | 3 | 2026-08-24 | |
| 13 | GPT-5.5 | OpenAI | 55.6 | 1 | 2026-08-23 |
| 14 | GPT-5.4 | OpenAI | 54.6 | 6 | 2026-08-23 |
| 15 | Inkling-Small | Thinking Machines | 54.4 | 1 | 2026-08-23 |
| 16 | Claude Sonnet 5 | Anthropic | 54.3 | 2 | 2026-08-23 |
| 17 | GPT-5.6 Luna | OpenAI | 53.4 | 1 | 2026-08-23 |
| 18 | GPT-5.6 Terra | OpenAI | 53.1 | 1 | 2026-08-23 |
| 19 | DeepSeek-V4-Pro | DeepSeek | 51.8 | 10 | 2026-08-23 |
| 20 | Kimi K2.6 | Moonshot | 50 | 4 | 2026-08-23 |
| 21 | laguna-s-2.1 | poolside | 49.7 | 1 | 2026-08-23 |
| 22 | Gemini 3 Flash Preview | 49.4 | 2 | 2026-08-23 | |
| 23 | Claude Sonnet 4.6 | Anthropic | 49.4 | 1 | 2026-06-30 |
| 24 | Gemini 3.1 Pro | 48.8 | 6 | 2026-07-04 | |
| 25 | Hy3-preview | Tencent | 48.5 | 1 | 2026-08-23 |
| 26 | GLM-5.2 Full Open Source | Z.ai | 48.2 | 1 | 2026-08-23 |
| 27 | MiniMax M2.7 | MiniMax | 46.3 | 1 | 2026-08-23 |
| 28 | GPT-5.2 | OpenAI | 46.3 | 2 | 2026-08-23 |
| 29 | MiniMax M2.1 | MiniMax | 43.5 | 2 | 2026-08-23 |
| 30 | Claude Opus 4.5 | Anthropic | 43.5 | 1 | 2026-06-05 |
| 31 | GPT-5.4 mini | OpenAI | 42.9 | 1 | 2026-08-23 |
| 32 | Claude Sonnet 4.5 | Anthropic | 41 | 2 | 2026-06-30 |
| 33 | Claudes | Anthropic | 41 | 1 | 2026-08-12 |
| 34 | GLM-5.1 | Z.ai | 40.7 | 3 | 2026-08-23 |
| 35 | Qwen3.6 Plus | Alibaba | 39.8 | 1 | 2026-08-23 |
| 36 | Qwen3.5 397B A17B | Alibaba | 38.3 | 1 | 2026-08-23 |
| 37 | Gemini 3 Pro | 36.4 | 1 | 2026-06-05 | |
| 38 | GPT-5.4 nano | OpenAI | 35.5 | 1 | 2026-08-23 |
| 39 | DeepSeek-V3.2 | DeepSeek | 35.2 | 2 | 2026-08-23 |
| 40 | DeepSeek-V3.2-Speciale | DeepSeek | 35.2 | 1 | 2026-08-23 |
| 41 | Kimi K2.5 | Moonshot | 27.8 | 1 | 2026-05-03 |
| 42 | Qwen3.6 35B A3B | Alibaba | 26.9 | 1 | 2026-08-23 |
| 43 | MiniMax M2 | MiniMax | 16.7 | 1 | 2026-06-05 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.