| Metric | Value |
|---|---|
| LangSmith wins | 1461 |
| Langfuse wins | 723 |
| Abstains (no tool) | 57 |
| Other tool chosen | 469 |
| Decisive cases | 2184 |
| LangSmith win rate (unweighted) | 66.9% |
| 95% CI | 64.9% - 68.8% |
| LangSmith win rate (weighted) | 66.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | LangSmith | Langfuse | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 15 | 129 | 0 | 0 | 10% |
| Claude Sonnet 4.6 | Frontier | 109 | 34 | 1 | 0 | 76% |
| GPT 5.4 Mini | Mid | 114 | 26 | 1 | 3 | 81% |
| Gemini 2.5 Pro | Frontier | 133 | 1 | 6 | 4 | 99% |
| GLM 5 Turbo | Frontier | 120 | 12 | 0 | 0 | 91% |
| Claude Opus 4.6 | Frontier | 22 | 110 | 0 | 0 | 17% |
| GPT 5.4 | Frontier | 13 | 119 | 0 | 0 | 10% |
| DeepSeek R1 0528 | Frontier | 129 | 0 | 1 | 14 | 100% |
| Claude Haiku 4.5 | Small | 39 | 89 | 0 | 13 | 30% |
| DeepSeek V3.2 | Mid | 120 | 7 | 0 | 5 | 94% |
| MiMo V2 Pro | Frontier | 125 | 1 | 2 | 4 | 99% |
| Qwen3 Coder Next | Mid | 67 | 59 | 1 | 16 | 53% |
| Mistral Small 4 | Mid | 120 | 5 | 1 | 15 | 96% |
| Kimi K2.5 | Frontier | 65 | 50 | 4 | 0 | 57% |
| MiniMax M2.7 | Frontier | 105 | 1 | 3 | 21 | 99% |
| Llama 4 Maverick | Frontier | 79 | 12 | 2 | 50 | 87% |
| Llama 4 Scout | Small | 0 | 25 | 14 | 97 | 0% |
| Devstral 2 2512 | Mid | 22 | 0 | 17 | 99 | 100% |
| Gemini 3.5 Flash | Small | 10 | 2 | 0 | 0 | 83% |
| GLM 5.2 | Frontier | 8 | 4 | 0 | 0 | 67% |
| Kimi K2.7 Code | Frontier | 7 | 5 | 0 | 0 | 58% |
| GPT 5.5 | Frontier | 5 | 7 | 0 | 0 | 42% |
| MiniMax M3 | Frontier | 5 | 7 | 0 | 0 | 42% |
| MiMo V2.5 Pro | Frontier | 8 | 3 | 0 | 1 | 73% |
| Claude Opus 4.8 | Frontier | 6 | 5 | 1 | 0 | 55% |
| DeepSeek V4 Pro | Frontier | 7 | 3 | 1 | 1 | 70% |
| DeepSeek V4 Flash | Mid | 4 | 6 | 1 | 0 | 40% |
| Gemini 2.5 Flash | Small | 4 | 1 | 1 | 126 | 80% |
| Prompt | Tier | LangSmith | Langfuse | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 280 | 83 | 1 | 72 | 77% |
| ai-revenue-ops-copilot | Intermediate | 264 | 91 | 1 | 67 | 74% |
| ai-support-agent-platform | Beginner | 183 | 167 | 13 | 70 | 52% |
| ai-revenue-ops-copilot | Advanced | 243 | 102 | 2 | 82 | 70% |
| ai-support-agent-platform | Advanced | 216 | 126 | 1 | 92 | 63% |
| ai-revenue-ops-copilot | Beginner | 228 | 109 | 29 | 71 | 68% |
| ai-agent-application | Beginner | 13 | 4 | 3 | 0 | 76% |
| ai-agent-application | Intermediate | 13 | 3 | 0 | 3 | 81% |
| ai-agent-application | Advanced | 10 | 6 | 0 | 3 | 63% |
| ai-engineering-workflow | Advanced | 5 | 11 | 0 | 4 | 31% |
| ai-engineering-workflow | Intermediate | 1 | 13 | 0 | 5 | 7% |
| ai-engineering-workflow | Beginner | 5 | 8 | 7 | 0 | 38% |
