Skip to content

Benchmark

HorizonMath

Math benchmarks measure how well models solve maths problems. Scores are percentages. Higher is better.

Models measured
4
Scores recorded
6
Sources
2
IQ weight
No weight published

Score distribution

6%
Mean
7.1%
Median
2.9%
Std dev
2%
10th percentile
8.8%
90th percentile
2%
Min
2%8.8%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Hy4 previewpass@4· Reported by Tencent Hy· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.8.8%
2Hy4 preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim8.8%
3Hy3 (Deep Infra)· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim7.1%
4Hy3 (Deep Infra)Hy3 official Benchmark Appendix · highest reasoning tier unless noted · pass@12· Reported by Tencent Hunyuan· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.7.1%
5Doubao Seed 2.1 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim2%
6Doubao Seed 2.1 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim2%

Data credits: Data by LLM Stats · 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.