FrontierCode 1.1 Extended - Pass Rate
Code · 2026-07-07
FrontierCode 1.1 Extended pass rate across the full set of 150 production-code tasks. A trial passes only when it satisfies every blocker criterion and is not flagged for unfair internet use. The result averages repeated trials per task.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 70.9 |
| Claude Opus 5.5 | 70.8 |
| GPT-6 Astra | 70.6 |
| Claude Sonnet 5.5 | 69.9 |
| Claude Opus 5 | 69.6 |
| Claude Fable 5.1 | 68.8 |
| SWE-2 | 68.4 |
| Grok 4.6 | 67.0 |
| GPT-5.6 Sol | 66.6 |
| GPT-6.1 Sol | 66.5 |
| GPT-6 Sol | 66.3 |
| Gemini 4 Argon | 66.1 |
| Opus 4.8 | 65.5 |
| Grok 4.7 | 65.2 |
| Claude Haiku 5.5 | 64.7 |
| GPT-5.5 | 62.8 |
| Gemini 3.7 Flash | 62.3 |
| Grok 4.5 | 62.3 |
| GPT-6 Luna | 62.3 |
| DeepSeek V4.1 Flash | 62.0 |
| GPT-5.6 Terra | 61.8 |
| Sonnet 5 | 61.7 |
| GPT-5.6 Luna | 60.9 |
| SWE-1.7 | 60.7 |
| Gemini 3.8 Flash | 59.6 |
| Opus 4.7 | 59.1 |
| Mistral Large 4 Preview | 58.1 |
| GLM-5.3 | 57.7 |
| Gemini 3.6 Flash | 53.3 |
| GLM-5.3 Flash | 51.1 |
| Kimi K2.7 Code | 50.0 |
| DeepSeek V4 Pro 0813 | 48.9 |
| Opus 4.6 | 48.0 |
| GPT-5.4 Mini | 47.9 |
| Composer 2.5 | 45.1 |
| Sonnet 4.6 | 44.4 |
| GLM-5.2 | 44.1 |
| DeepSeek-V4-Flash-0731 | 35.0 |
| DeepSeek-V4-Pro | 34.5 |
| MiniMax M3 | 31.5 |