Toolathlon
Can the model plan, use tools, and finish a multi-step task?
Agents11 ranked models12 reported values2 reportsHigher is better
Top rankings
One row per model, using its best reported score across effort settings.
| Rank | Model | Best score | Best reported setting |
|---|---|---|---|
| 1 | Claude Fable 5Anthropic | 61.7% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 2 | Claude Mythos 5Anthropic | 61.7% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 3 | Claude Mythos PreviewAnthropic | 61.1% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 4 | Claude Opus 4.8Anthropic | 59.9% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 5 | GPT-5.6 SolOpenAI | 58.0% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 6 | Gemini 3.5 FlashGoogle | 56.5% | Reported configurationGoogle report1 value · 1 reportMay 19, 2026 · source |
| 7 | GPT-5.5OpenAI | 55.6% | Reported configurationOpenAI report2 values · 2 reportsJul 9, 2026 · source |
| 8 | GPT-5.6 LunaOpenAI | 53.4% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 9 | GPT-5.6 TerraOpenAI | 53.1% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
| 10 | Gemini 3 FlashGoogle | 49.4% | Reported configurationGoogle report1 value · 1 reportMay 19, 2026 · source |
| 11 | Gemini 3.1 Pro PreviewGoogle | 48.8% | Reported configurationOpenAI report1 value · 1 reportJul 9, 2026 · source |
Effort curve
Every sourced cost-linked effort value for this exact version. Lines connect complete sweeps only.
No cost-linked effort sweep for this version.
Definition and comparison boundarypublic methodology
Can the model plan, use tools, and finish a multi-step task?
Agentic execution across tools, environments, or long-running workflows. Higher is better. The value is the percentage reported in this lab's table.
Harness, tools, prompts, attempt count, and benchmark version can materially change the result. Compare within one reporting context.
Method / source