| Metric | Value |
|---|---|
| Python wins | 693 |
| TypeScript wins | 238 |
| Abstains (no tool) | 26 |
| Other tool chosen | 109 |
| Decisive cases | 931 |
| Python win rate (unweighted) | 74.4% |
| 95% CI | 71.5% - 77.1% |
| Python win rate (weighted) | 74.4% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | Python | TypeScript | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 35 | 9 | 0 | 9 | 80% |
| DeepSeek Pro 4–4 0813 | Frontier | 34 | 15 | 3 | 2 | 69% |
| DeepSeek R1 | Frontier | 41 | 2 | 5 | 6 | 95% |
| Devstral | Mid | 34 | 8 | 0 | 0 | 81% |
| Gemini Flash 2.5–3.8 | Small | 35 | 18 | 0 | 1 | 66% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 33 | 16 | 1 | 4 | 67% |
| GLM 5 Turbo–5.3 | Frontier | 30 | 22 | 0 | 2 | 58% |
| GPT 5.4–6 Astra | Frontier | 37 | 14 | 0 | 3 | 73% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 42 | 11 | 1 | 0 | 79% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 32 | 19 | 0 | 3 | 63% |
| Haiku | Small | 44 | 1 | 0 | 9 | 98% |
| Kimi 2.5–3 | Frontier | 26 | 22 | 4 | 2 | 54% |
| Llama Maverick | Frontier | 26 | 6 | 6 | 16 | 81% |
| Llama Scout | Small | 42 | 0 | 0 | 12 | 100% |
| MiMo Pro 2–2.6 | Frontier | 43 | 9 | 0 | 2 | 83% |
| MiniMax M 2.7–3 | Frontier | 29 | 21 | 0 | 4 | 58% |
| Mistral Small | Mid | 26 | 1 | 0 | 26 | 96% |
| Opus 4.6–5.5 | Frontier | 39 | 15 | 0 | 0 | 72% |
| Qwen Coder | Mid | 25 | 16 | 6 | 7 | 61% |
| Sonnet 4.6–5.5 | Frontier | 40 | 13 | 0 | 1 | 75% |
| Prompt | Tier | Python | TypeScript | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-agent-application | Beginner | 174 | 1 | 2 | 1 | 99% |
| ai-agent-application | Intermediate | 170 | 1 | 7 | 0 | 99% |
| ai-agent-application | Advanced | 163 | 1 | 12 | 0 | 99% |
| full-stack-web-app | Intermediate | 55 | 94 | 0 | 29 | 37% |
| full-stack-web-app | Beginner | 102 | 39 | 4 | 33 | 72% |
| full-stack-web-app | Advanced | 29 | 102 | 1 | 46 | 22% |
