| Metric | Value |
|---|---|
| Ragas wins | 116 |
| Braintrust wins | 806 |
| Abstains (no tool) | 105 |
| Other tool chosen | 1648 |
| Decisive cases | 922 |
| Ragas win rate (unweighted) | 12.6% |
| 95% CI | 10.6% - 14.9% |
| Ragas win rate (weighted) | 12.6% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Ragas | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 0 | 136 | 0 | 8 | 0% |
| Claude Opus 4.6 | Frontier | 18 | 113 | 0 | 1 | 14% |
| Claude Haiku 4.5 | Small | 0 | 113 | 1 | 23 | 0% |
| Kimi K2.5 | Frontier | 1 | 109 | 3 | 6 | 1% |
| GPT 5.4 | Frontier | 1 | 92 | 0 | 39 | 1% |
| GLM 5 Turbo | Frontier | 6 | 84 | 19 | 23 | 7% |
| Claude Sonnet 4.6 | Frontier | 7 | 65 | 1 | 71 | 10% |
| MiniMax M2.7 | Frontier | 24 | 31 | 5 | 69 | 44% |
| MiMo V2 Pro | Frontier | 24 | 1 | 8 | 99 | 96% |
| GPT 5.4 Mini | Mid | 9 | 4 | 3 | 127 | 69% |
| GPT 5.5 | Frontier | 0 | 12 | 0 | 0 | 0% |
| GLM 5.2 | Frontier | 2 | 9 | 0 | 1 | 18% |
| MiniMax M3 | Frontier | 1 | 10 | 1 | 0 | 9% |
| Mistral Small 4 | Mid | 9 | 0 | 2 | 125 | 100% |
| Kimi K2.7 Code | Frontier | 3 | 6 | 1 | 2 | 33% |
| Gemini 3.5 Flash | Small | 0 | 9 | 1 | 2 | 0% |
| MiMo V2.5 Pro | Frontier | 2 | 5 | 0 | 5 | 29% |
| DeepSeek V4 Pro | Frontier | 0 | 4 | 1 | 7 | 0% |
| Llama 4 Scout | Small | 3 | 0 | 7 | 121 | 100% |
| Claude Opus 4.8 | Frontier | 2 | 1 | 1 | 8 | 67% |
| DeepSeek V3.2 | Mid | 2 | 0 | 22 | 104 | 100% |
| DeepSeek V4 Flash | Mid | 1 | 1 | 1 | 9 | 50% |
| Devstral 2 2512 | Mid | 1 | 0 | 6 | 128 | 100% |
| Gemini 2.5 Pro | Frontier | 0 | 1 | 9 | 134 | 0% |
| DeepSeek R1 0528 | Frontier | 0 | 0 | 7 | 137 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 1 | 126 | n/a |
| Llama 4 Maverick | Frontier | 0 | 0 | 2 | 135 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 3 | 138 | n/a |
| Prompt | Tier | Ragas | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Advanced | 9 | 164 | 2 | 244 | 5% |
| ai-support-agent-platform | Advanced | 37 | 130 | 5 | 258 | 22% |
| ai-revenue-ops-copilot | Beginner | 15 | 151 | 10 | 254 | 9% |
| ai-support-agent-platform | Beginner | 34 | 100 | 66 | 231 | 25% |
| ai-revenue-ops-copilot | Intermediate | 2 | 130 | 4 | 288 | 2% |
| ai-support-agent-platform | Intermediate | 8 | 92 | 5 | 323 | 8% |
| ai-engineering-workflow | Advanced | 0 | 12 | 0 | 6 | 0% |
| ai-agent-application | Intermediate | 5 | 6 | 0 | 9 | 45% |
| ai-agent-application | Advanced | 3 | 7 | 0 | 8 | 30% |
| ai-engineering-workflow | Intermediate | 0 | 8 | 0 | 9 | 0% |
| ai-agent-application | Beginner | 3 | 3 | 5 | 9 | 50% |
| ai-engineering-workflow | Beginner | 0 | 3 | 8 | 9 | 0% |