Skip to content

Benchmark

MultiLangBench · 20-language average test pass rate

Scores are percentages. Higher is better.

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

Score distribution

51.5%
Mean
40.3%
Median
26%
Std dev
26.8%
10th percentile
87.4%
90th percentile
26.8%
Min
26.8%87.4%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1GPT-6 AstraOpenAI publication · SQLite 3.40.1 from scratch · 20-language average · maximum reasoning effort · 272-test recommended subset in public harness; exact publication test selection and API snapshot not stated· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.87.4%
2GPT-6 LunaOpenAI publication · SQLite 3.40.1 from scratch · 20-language average · maximum reasoning effort · 272-test recommended subset in public harness; exact publication test selection and API snapshot not stated· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.40.3%
3GPT-5.6 SolOpenAI publication · SQLite 3.40.1 from scratch · 20-language average · maximum reasoning effort · 272-test recommended subset in public harness; exact publication test selection and API snapshot not stated· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.26.8%

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.