| Metric | Value |
|---|---|
| Ragas wins | 47 |
| Arize AI wins | 47 |
| Abstains (no tool) | 36 |
| Other tool chosen | 870 |
| Decisive cases | 94 |
| Ragas win rate (unweighted) | 50.0% |
| 95% CI | 40.1% - 59.9% |
| Ragas win rate (weighted) | 50.0% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Ragas | Arize AI | None | Other | A rate |
|---|---|---|---|---|---|---|
| Llama 4 Scout | Small | 1 | 19 | 3 | 26 | 5% |
| MiMo V2 Pro | Frontier | 13 | 0 | 2 | 39 | 100% |
| Gemini 2.5 Flash | Small | 0 | 12 | 0 | 40 | 0% |
| MiniMax M2.7 | Frontier | 10 | 0 | 1 | 41 | 100% |
| Claude Opus 4.6 | Frontier | 8 | 0 | 0 | 46 | 100% |
| GPT 5.4 Mini | Mid | 5 | 0 | 1 | 48 | 100% |
| Devstral 2 2512 | Mid | 1 | 4 | 1 | 45 | 20% |
| Gemini 2.5 Pro | Frontier | 0 | 5 | 6 | 43 | 0% |
| Llama 4 Maverick | Frontier | 0 | 5 | 0 | 48 | 0% |
| Mistral Small 4 | Mid | 4 | 0 | 0 | 47 | 100% |
| Claude Sonnet 4.6 | Frontier | 1 | 0 | 0 | 53 | 100% |
| DeepSeek V3.2 | Mid | 1 | 0 | 9 | 42 | 100% |
| GLM 5 Turbo | Frontier | 1 | 0 | 7 | 46 | 100% |
| GPT 5.4 | Frontier | 1 | 0 | 0 | 53 | 100% |
| Kimi K2.5 | Frontier | 1 | 0 | 0 | 47 | 100% |
| DeepSeek R1 0528 | Frontier | 0 | 1 | 2 | 51 | 0% |
| Qwen3 Coder Next | Mid | 0 | 1 | 3 | 50 | 0% |
| Claude Haiku 4.5 | Small | 0 | 0 | 1 | 51 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 54 | n/a |
| Prompt | Tier | Ragas | Arize AI | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Beginner | 17 | 7 | 25 | 120 | 71% |
| ai-support-agent-platform | Advanced | 13 | 11 | 1 | 144 | 54% |
| ai-revenue-ops-copilot | Advanced | 6 | 10 | 1 | 147 | 38% |
| ai-revenue-ops-copilot | Beginner | 6 | 5 | 4 | 153 | 55% |
| ai-support-agent-platform | Intermediate | 3 | 8 | 4 | 151 | 27% |
| ai-revenue-ops-copilot | Intermediate | 2 | 6 | 1 | 155 | 25% |