| Metric | Value |
|---|---|
| LangSmith wins | 2417 |
| Langfuse wins | 1365 |
| Abstains (no tool) | 159 |
| Other tool chosen | 648 |
| Decisive cases | 3782 |
| LangSmith win rate (unweighted) | 63.9% |
| 95% CI | 62.4% - 65.4% |
| LangSmith win rate (weighted) | 63.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | LangSmith | Langfuse | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 181 | 48 | 4 | 6 | 79% |
| DeepSeek Pro 4–4 0813 | Frontier | 68 | 37 | 1 | 2 | 65% |
| DeepSeek R1 | Frontier | 214 | 1 | 2 | 23 | 100% |
| Devstral | Mid | 39 | 0 | 29 | 141 | 100% |
| Gemini Flash 2.5–3.8 | Small | 81 | 25 | 4 | 130 | 76% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 217 | 9 | 9 | 5 | 96% |
| GLM 5 Turbo–5.3 | Frontier | 186 | 54 | 0 | 0 | 78% |
| GPT 5.4–6 Astra | Frontier | 58 | 176 | 5 | 1 | 25% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 30 | 208 | 2 | 0 | 13% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 165 | 65 | 6 | 4 | 72% |
| Haiku | Small | 59 | 153 | 3 | 21 | 28% |
| Kimi 2.5–3 | Frontier | 115 | 106 | 4 | 1 | 52% |
| Llama Maverick | Frontier | 130 | 23 | 17 | 68 | 85% |
| Llama Scout | Small | 0 | 37 | 36 | 149 | 0% |
| MiMo Pro 2–2.6 | Frontier | 207 | 19 | 8 | 6 | 92% |
| MiniMax M 2.7–3 | Frontier | 153 | 57 | 5 | 22 | 73% |
| Mistral Small | Mid | 181 | 11 | 4 | 39 | 94% |
| Opus 4.6–5.5 | Frontier | 64 | 167 | 9 | 0 | 28% |
| Qwen Coder | Mid | 117 | 91 | 2 | 29 | 56% |
| Sonnet 4.6–5.5 | Frontier | 152 | 78 | 9 | 1 | 66% |
| Prompt | Tier | LangSmith | Langfuse | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 391 | 117 | 1 | 85 | 77% |
| ai-revenue-ops-copilot | Intermediate | 362 | 136 | 1 | 79 | 73% |
| ai-support-agent-platform | Beginner | 263 | 227 | 17 | 82 | 54% |
| ai-revenue-ops-copilot | Advanced | 341 | 144 | 3 | 97 | 70% |
| ai-support-agent-platform | Advanced | 303 | 178 | 1 | 110 | 63% |
| ai-revenue-ops-copilot | Beginner | 315 | 155 | 34 | 91 | 67% |
| ai-agent-application | Advanced | 94 | 68 | 0 | 13 | 58% |
| ai-agent-application | Intermediate | 115 | 45 | 1 | 14 | 72% |
| ai-engineering-workflow | Intermediate | 50 | 100 | 0 | 27 | 33% |
| ai-agent-application | Beginner | 110 | 37 | 27 | 4 | 75% |
| ai-engineering-workflow | Advanced | 48 | 91 | 0 | 34 | 35% |
| ai-engineering-workflow | Beginner | 25 | 67 | 74 | 12 | 27% |
