Leading: Playwright (66.9%)
| Metric | Value |
|---|---|
| Playwright wins | 232 |
| Browserbase wins | 115 |
| Abstains (no tool) | 148 |
| Other tool chosen | 34 |
| Decisive cases | 347 |
| Playwright win rate (unweighted) | 66.9% |
| 95% CI | 61.7% - 71.6% |
| Playwright win rate (weighted) | 66.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | Playwright | Browserbase | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 21 | 0 | 6 | 0 | 100% |
| DeepSeek Pro 4–4 0813 | Frontier | 15 | 4 | 8 | 0 | 79% |
| DeepSeek R1 | Frontier | 18 | 0 | 9 | 0 | 100% |
| Devstral | Mid | 13 | 0 | 7 | 1 | 100% |
| Gemini Flash 2.5–3.8 | Small | 5 | 14 | 7 | 1 | 26% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 6 | 3 | 8 | 6 | 67% |
| GLM 5 Turbo–5.3 | Frontier | 5 | 13 | 8 | 1 | 28% |
| GPT 5.4–6 Astra | Frontier | 2 | 17 | 8 | 0 | 11% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 15 | 5 | 7 | 0 | 75% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 20 | 4 | 3 | 0 | 83% |
| Haiku | Small | 13 | 5 | 9 | 0 | 72% |
| Kimi 2.5–3 | Frontier | 5 | 14 | 8 | 0 | 26% |
| Llama Maverick | Frontier | 18 | 0 | 9 | 0 | 100% |
| Llama Scout | Small | 7 | 0 | 9 | 11 | 100% |
| MiMo Pro 2–2.6 | Frontier | 10 | 8 | 8 | 1 | 56% |
| MiniMax M 2.7–3 | Frontier | 3 | 19 | 2 | 2 | 14% |
| Mistral Small | Mid | 22 | 0 | 5 | 0 | 100% |
| Opus 4.6–5.5 | Frontier | 11 | 7 | 9 | 0 | 61% |
| Qwen Coder | Mid | 7 | 0 | 9 | 11 | 100% |
| Sonnet 4.6–5.5 | Frontier | 16 | 2 | 9 | 0 | 89% |
| Prompt | Tier | Playwright | Browserbase | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-agent-application | Intermediate | 105 | 56 | 1 | 16 | 65% |
| ai-agent-application | Advanced | 106 | 52 | 0 | 15 | 67% |
| ai-agent-application | Beginner | 21 | 7 | 147 | 3 | 75% |
