Skip to content

Benchmark

MathVision · Without CI

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

Models measured
10
Scores recorded
11
Sources
1
IQ weight
No weight published

Score distribution

88.2%
Mean
90.3%
Median
7.7%
Std dev
84.2%
10th percentile
94.3%
90th percentile
65.5%
Min
65.5%95.2%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Qwen3.8 MaxQwen3.8-Max release vision/video table · without Code Interpreter· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.95.2%
2Kimi K3no tools · three-run average · max effort · temp 1.0 · top_p .95· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.94.3%
3Qwen3.8 Omni FlashWithout Code Interpreter · Qwen 固定 boxed prompt;其他模型取有/无 boxed 较高值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.91.8%
4Qwen 3.8 Flash NextQwen3.8-Flash-Next release/model card · without Code Interpreter· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90.6%
5Qwen3.8 FlashWithout Code Interpreter · Qwen 固定 boxed prompt;其他模型取有/无 boxed 较高值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90.6%
6Qwen3.7 PlusWithout Code Interpreter · Qwen 固定 boxed prompt;其他模型取有/无 boxed 较高值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90.3%
7Qwen 3.8 27BQwen3.8-27B model card · without Code Interpreter· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90%
8Qwen 3.8 27BWithout Code Interpreter · Qwen 固定 boxed prompt;其他模型取有/无 boxed 较高值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90%
9Kimi K2.6Kimi K2.6 official benchmark table · thinking · temp 1.0 · top_p 1.0 · 262,144 context · no Python · avg@3· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.87.4%
10Kimi K2.5official model-card base evaluation· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.84.2%
11Claude Opus 4.6Without Code Interpreter · Qwen 固定 boxed prompt;其他模型取有/无 boxed 较高值 · max effort· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.65.5%

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.