| Metric | Value |
|---|---|
| Langfuse wins | 80 |
| Promptfoo wins | 90 |
| Abstains (no tool) | 183 |
| Other tool chosen | 3469 |
| Decisive cases | 170 |
| Langfuse win rate (unweighted) | 47.1% |
| 95% CI | 39.7% - 54.5% |
| Langfuse win rate (weighted) | 47.1% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Langfuse | Promptfoo | None | Other | A rate |
|---|---|---|---|---|---|---|
| Qwen3 Coder Next | Mid | 23 | 0 | 4 | 171 | 100% |
| GPT 5.4 Mini | Mid | 9 | 9 | 8 | 175 | 50% |
| Claude Sonnet 4.6 | Frontier | 16 | 0 | 6 | 182 | 100% |
| GLM 5.2 | Frontier | 0 | 14 | 0 | 58 | 0% |
| Claude Opus 4.8 | Frontier | 11 | 1 | 6 | 53 | 92% |
| Mistral Small 4 | Mid | 1 | 10 | 3 | 175 | 9% |
| Gemini 3.5 Flash | Small | 0 | 11 | 7 | 54 | 0% |
| Kimi K2.7 Code | Frontier | 0 | 11 | 1 | 59 | 0% |
| MiMo V2.5 Pro | Frontier | 2 | 8 | 8 | 53 | 20% |
| DeepSeek V4 Flash | Mid | 3 | 4 | 4 | 60 | 43% |
| Llama 4 Scout | Small | 6 | 0 | 20 | 159 | 100% |
| MiMo V2 Pro | Frontier | 0 | 6 | 8 | 118 | 0% |
| GLM 5 Turbo | Frontier | 0 | 5 | 19 | 108 | 0% |
| Claude Haiku 4.5 | Small | 4 | 0 | 3 | 187 | 100% |
| Gemini 2.5 Pro | Frontier | 4 | 0 | 11 | 187 | 100% |
| DeepSeek V4 Pro | Frontier | 0 | 4 | 2 | 66 | 0% |
| MiniMax M3 | Frontier | 0 | 3 | 5 | 64 | 0% |
| Kimi K2.5 | Frontier | 0 | 2 | 3 | 114 | 0% |
| DeepSeek V3.2 | Mid | 1 | 0 | 22 | 105 | 100% |
| GPT 5.3 Codex | Frontier | 0 | 1 | 0 | 203 | 0% |
| GPT 5.5 | Frontier | 0 | 1 | 2 | 69 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 0 | 0 | 132 | n/a |
| DeepSeek R1 0528 | Frontier | 0 | 0 | 9 | 195 | n/a |
| Devstral 2 2512 | Mid | 0 | 0 | 14 | 163 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 1 | 126 | n/a |
| GPT 5.4 | Frontier | 0 | 0 | 0 | 132 | n/a |
| Llama 4 Maverick | Frontier | 0 | 0 | 12 | 177 | n/a |
| MiniMax M2.7 | Frontier | 0 | 0 | 5 | 124 | n/a |
| Prompt | Tier | Langfuse | Promptfoo | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Beginner | 30 | 4 | 73 | 420 | 88% |
| ai-engineering-workflow | Beginner | 5 | 18 | 53 | 40 | 22% |
| ai-revenue-ops-copilot | Intermediate | 10 | 11 | 4 | 497 | 48% |
| ai-engineering-workflow | Intermediate | 7 | 13 | 1 | 91 | 35% |
| ai-engineering-workflow | Advanced | 3 | 16 | 1 | 92 | 16% |
| ai-support-agent-platform | Intermediate | 9 | 5 | 6 | 502 | 64% |
| ai-revenue-ops-copilot | Beginner | 6 | 5 | 11 | 504 | 55% |
| ai-support-agent-platform | Advanced | 6 | 5 | 5 | 511 | 55% |
| ai-revenue-ops-copilot | Advanced | 0 | 7 | 3 | 504 | 0% |
| ai-agent-application | Beginner | 4 | 2 | 25 | 88 | 67% |
| ai-agent-application | Intermediate | 0 | 3 | 1 | 109 | 0% |
| ai-agent-application | Advanced | 0 | 1 | 0 | 111 | 0% |
