| Metric | Value |
|---|---|
| Cursor wins | 1219 |
| LangChain wins | 358 |
| Abstains (no tool) | 855 |
| Other tool chosen | 873 |
| Decisive cases | 1577 |
| Cursor win rate (unweighted) | 77.3% |
| 95% CI | 75.2% - 79.3% |
| Cursor win rate (weighted) | 77.3% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Cursor | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.4 Mini | Mid | 135 | 1 | 8 | 31 | 99% |
| GLM 5 Turbo | Frontier | 130 | 0 | 2 | 0 | 100% |
| GPT 5.3 Codex | Frontier | 123 | 0 | 27 | 24 | 100% |
| Claude Opus 4.6 | Frontier | 74 | 38 | 20 | 0 | 66% |
| Claude Sonnet 4.6 | Frontier | 107 | 0 | 70 | 0 | 100% |
| GPT 5.4 | Frontier | 87 | 0 | 31 | 10 | 100% |
| DeepSeek V3.2 | Mid | 57 | 17 | 52 | 4 | 77% |
| MiniMax M2.7 | Frontier | 54 | 10 | 25 | 40 | 84% |
| Llama 4 Maverick | Frontier | 14 | 50 | 64 | 48 | 22% |
| MiMo V2 Pro | Frontier | 4 | 57 | 30 | 41 | 7% |
| DeepSeek R1 0528 | Frontier | 1 | 58 | 62 | 53 | 2% |
| Kimi K2.5 | Frontier | 56 | 0 | 53 | 9 | 100% |
| Mistral Small 4 | Mid | 48 | 3 | 66 | 36 | 94% |
| Qwen3 Coder Next | Mid | 12 | 39 | 20 | 102 | 24% |
| Gemini 2.5 Flash | Small | 2 | 49 | 59 | 22 | 4% |
| GLM 5.2 | Frontier | 45 | 0 | 0 | 0 | 100% |
| MiniMax M3 | Frontier | 39 | 0 | 3 | 3 | 100% |
| Claude Opus 4.8 | Frontier | 35 | 0 | 6 | 4 | 100% |
| GPT 5.5 | Frontier | 35 | 0 | 2 | 8 | 100% |
| Gemini 3.5 Flash | Small | 30 | 0 | 8 | 1 | 100% |
| Kimi K2.7 Code | Frontier | 25 | 0 | 5 | 14 | 100% |
| MiMo V2.5 Pro | Frontier | 24 | 1 | 8 | 12 | 96% |
| DeepSeek V4 Pro | Frontier | 22 | 3 | 8 | 12 | 88% |
| Gemini 2.5 Pro | Frontier | 21 | 0 | 22 | 130 | 100% |
| Devstral 2 2512 | Mid | 6 | 15 | 64 | 85 | 29% |
| DeepSeek V4 Flash | Mid | 15 | 4 | 10 | 15 | 79% |
| Claude Haiku 4.5 | Small | 18 | 0 | 79 | 67 | 100% |
| Llama 4 Scout | Small | 0 | 13 | 51 | 102 | 0% |
| Prompt | Tier | Cursor | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Advanced | 231 | 76 | 8 | 171 | 75% |
| ai-revenue-ops-copilot | Intermediate | 247 | 45 | 2 | 205 | 85% |
| ai-support-agent-platform | Advanced | 247 | 41 | 0 | 215 | 86% |
| ai-support-agent-platform | Intermediate | 89 | 90 | 243 | 92 | 50% |
| ai-revenue-ops-copilot | Beginner | 111 | 58 | 272 | 73 | 66% |
| ai-support-agent-platform | Beginner | 92 | 47 | 330 | 48 | 66% |
| ai-engineering-workflow | Beginner | 76 | 0 | 0 | 20 | 100% |
| ai-engineering-workflow | Intermediate | 74 | 0 | 0 | 23 | 100% |
| ai-engineering-workflow | Advanced | 52 | 1 | 0 | 26 | 98% |
