Skip to content

Benchmark

BankerToolBench

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

Models measured
3
Scores recorded
7
Sources
2
IQ weight
No weight published

Score distribution

76.8%
Mean
76.2%
Median
1.1%
Std dev
76.1%
10th percentile
78.6%
90th percentile
76.1%
Min
76.1%78.6%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Hy4 preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim78.6%
2Hy4 previewofficial model-card appendix· Reported by Tencent Hy· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.78.6%
3MiniMax M3.1 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim76.2%
4MiniMax M3.1 FlashMiniMax M3.1 Flash official-weights model-card run · reasoning effort, sampling, and run count not stated for this benchmark · OpenCode 1.17.18 · MiniMax M2.7 judge· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.76.2%
5MiniMax M3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim76.1%
6MiniMax M3Claude Code · MiniMax M2.7 grader· Reported by MiniMax Research· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.76.1%
7MiniMax M3official model-page overview figure · rounded· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.76.1%

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. · Data by LLM Stats