| Metric | Value |
|---|---|
| Langfuse wins | 723 |
| LangSmith wins | 1461 |
| Abstains (no tool) | 57 |
| Other tool chosen | 469 |
| Decisive cases | 2184 |
| Langfuse win rate (unweighted) | 33.1% |
| 95% CI | 31.2% - 35.1% |
| Langfuse win rate (weighted) | 33.1% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Langfuse | LangSmith | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 129 | 15 | 0 | 0 | 90% |
| Claude Sonnet 4.6 | Frontier | 34 | 109 | 1 | 0 | 24% |
| GPT 5.4 Mini | Mid | 26 | 114 | 1 | 3 | 19% |
| Gemini 2.5 Pro | Frontier | 1 | 133 | 6 | 4 | 1% |
| GPT 5.4 | Frontier | 119 | 13 | 0 | 0 | 90% |
| Claude Opus 4.6 | Frontier | 110 | 22 | 0 | 0 | 83% |
| GLM 5 Turbo | Frontier | 12 | 120 | 0 | 0 | 9% |
| DeepSeek R1 0528 | Frontier | 0 | 129 | 1 | 14 | 0% |
| Claude Haiku 4.5 | Small | 89 | 39 | 0 | 13 | 70% |
| DeepSeek V3.2 | Mid | 7 | 120 | 0 | 5 | 6% |
| Qwen3 Coder Next | Mid | 59 | 67 | 1 | 16 | 47% |
| MiMo V2 Pro | Frontier | 1 | 125 | 2 | 4 | 1% |
| Mistral Small 4 | Mid | 5 | 120 | 1 | 15 | 4% |
| Kimi K2.5 | Frontier | 50 | 65 | 4 | 0 | 43% |
| MiniMax M2.7 | Frontier | 1 | 105 | 3 | 21 | 1% |
| Llama 4 Maverick | Frontier | 12 | 79 | 2 | 50 | 13% |
| Llama 4 Scout | Small | 25 | 0 | 14 | 97 | 100% |
| Devstral 2 2512 | Mid | 0 | 22 | 17 | 99 | 0% |
| GPT 5.5 | Frontier | 7 | 5 | 0 | 0 | 58% |
| MiniMax M3 | Frontier | 7 | 5 | 0 | 0 | 58% |
| Kimi K2.7 Code | Frontier | 5 | 7 | 0 | 0 | 42% |
| GLM 5.2 | Frontier | 4 | 8 | 0 | 0 | 33% |
| Gemini 3.5 Flash | Small | 2 | 10 | 0 | 0 | 17% |
| Claude Opus 4.8 | Frontier | 5 | 6 | 1 | 0 | 45% |
| MiMo V2.5 Pro | Frontier | 3 | 8 | 0 | 1 | 27% |
| DeepSeek V4 Flash | Mid | 6 | 4 | 1 | 0 | 60% |
| DeepSeek V4 Pro | Frontier | 3 | 7 | 1 | 1 | 30% |
| Gemini 2.5 Flash | Small | 1 | 4 | 1 | 126 | 20% |
| Prompt | Tier | Langfuse | LangSmith | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 83 | 280 | 1 | 72 | 23% |
| ai-revenue-ops-copilot | Intermediate | 91 | 264 | 1 | 67 | 26% |
| ai-support-agent-platform | Beginner | 167 | 183 | 13 | 70 | 48% |
| ai-revenue-ops-copilot | Advanced | 102 | 243 | 2 | 82 | 30% |
| ai-support-agent-platform | Advanced | 126 | 216 | 1 | 92 | 37% |
| ai-revenue-ops-copilot | Beginner | 109 | 228 | 29 | 71 | 32% |
| ai-agent-application | Beginner | 4 | 13 | 3 | 0 | 24% |
| ai-engineering-workflow | Advanced | 11 | 5 | 0 | 4 | 69% |
| ai-agent-application | Advanced | 6 | 10 | 0 | 3 | 38% |
| ai-agent-application | Intermediate | 3 | 13 | 0 | 3 | 19% |
| ai-engineering-workflow | Intermediate | 13 | 1 | 0 | 5 | 93% |
| ai-engineering-workflow | Beginner | 8 | 5 | 7 | 0 | 62% |
