MultiLangBench - Python: leaderboard
OpenAI October 19 release; implement SQLite 3.40.1 from documentation without network or installed SQLite; validate database effects against reference SQLite. Default released evaluation uses 272 hash-selected tests; the repository permits full-suite refinement if reported. Every embedded model/effort point retained; no cost-based or best-effort selection. Test Pass Rate (%); higher is better.
Metric: Test Pass Rate (%). Source: openai.com. Saturation forecast: Around January 2027. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6.1 Sol (Max) | 91 | #7 (GPT-6.1 Sol) |
| 2 | GPT-6 Astra (Max) | 90.1 | #6 (GPT-6 Astra) |
| 3 | GPT-6 Astra (xHigh) | 79 | #6 (GPT-6 Astra) |
| 4 | GPT-6 Astra (High) | 78.1 | #6 (GPT-6 Astra) |
| 5 | GPT-6.1 Sol (xHigh) | 76.8 | #7 (GPT-6.1 Sol) |
| 6 | GPT-6.1 Sol (High) | 71.7 | #7 (GPT-6.1 Sol) |
| 7 | GPT-6 Astra (Medium) | 65.7 | #6 (GPT-6 Astra) |
| 8 | GPT-6.1 Sol (Medium) | 62.2 | #7 (GPT-6.1 Sol) |
| 9 | GPT-6 Luna (Max) | 49.2 | #56 (GPT-6 Luna) |
| 10 | GPT-5.6 Sol (Max) | 34.2 | #19 (GPT-5.6 Sol) |
| 11 | GPT-6 Luna (xHigh) | 31.5 | #56 (GPT-6 Luna) |
| 12 | GPT-5.6 Sol (xHigh) | 23.3 | #19 (GPT-5.6 Sol) |
| 13 | GPT-6 Astra (Low) | 23 | #6 (GPT-6 Astra) |
| 14 | GPT-6 Luna (High) | 22.1 | #56 (GPT-6 Luna) |
| 15 | GPT-6.1 Sol (Low) | 12.9 | #7 (GPT-6.1 Sol) |
No result here: #3 Claude Opus 5.5, #8 Gemini 4 Argon, #9 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=multilangbench-python · How It Works · Data refreshed daily, snapshot 2026-10-20.