Skip to content

Benchmark

Toolathlon Verified · Pass³

Scores are percentages. Higher is better.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
No weight published

Score distribution

70.7%
Mean
72.2%
Median
2.7%
Std dev
68.5%
10th percentile
73.1%
90th percentile
65.7%
Min
65.7%73.1%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Claude Fable 5.1internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.73.1%
2Claude Opus 5internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.73.1%
3Claude Opus 5.5internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.72.2%
4Claude Opus 4.8internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.71.3%
5Claude Sonnet 5.5internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.68.5%
6Claude Sonnet 5internal pinned harness · adaptive thinking · max · 108 tasks · 3 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.65.7%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.