Skip to content

Benchmark

Legal Agent Full Public Benchmark

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.

Models measured
3
Scores recorded
5
Sources
1
IQ weight
No weight published

Score distribution

40.7%
Mean
9.6%
Median
39%
Std dev
8%
10th percentile
88.5%
90th percentile
8%
Min
8%88.5%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Claude Sonnet 4.61,235-task public set · mean criterion pass · internal reduced-tool harness · 5 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.88.5%
2Claude Sonnet 51,235-task public set · mean criterion pass · internal reduced-tool harness · 5 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.88.3%
3Claude Opus 4.8Fable/Mythos 5 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.9.6%
4Claude Sonnet 51,235-task public set · all-pass · internal reduced-tool harness · 5 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.8.9%
5Claude Sonnet 4.61,235-task public set · all-pass · internal reduced-tool harness · 5 trials· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.8%

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.