| Metric | Value |
|---|---|
| Arize AI wins | 108 |
| Ragas wins | 116 |
| Abstains (no tool) | 105 |
| Other tool chosen | 2346 |
| Decisive cases | 224 |
| Arize AI win rate (unweighted) | 48.2% |
| 95% CI | 41.8% - 54.7% |
| Arize AI win rate (weighted) | 48.2% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Arize AI | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| Llama 4 Scout | Small | 41 | 3 | 7 | 80 | 93% |
| Gemini 2.5 Flash | Small | 27 | 0 | 1 | 99 | 100% |
| MiniMax M2.7 | Frontier | 1 | 24 | 5 | 99 | 4% |
| MiMo V2 Pro | Frontier | 0 | 24 | 8 | 100 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 18 | 0 | 114 | 0% |
| Llama 4 Maverick | Frontier | 13 | 0 | 2 | 122 | 100% |
| Mistral Small 4 | Mid | 3 | 9 | 2 | 122 | 25% |
| Gemini 2.5 Pro | Frontier | 10 | 0 | 9 | 125 | 100% |
| Devstral 2 2512 | Mid | 9 | 1 | 6 | 119 | 90% |
| GPT 5.4 Mini | Mid | 0 | 9 | 3 | 131 | 0% |
| Claude Sonnet 4.6 | Frontier | 0 | 7 | 1 | 136 | 0% |
| GLM 5 Turbo | Frontier | 0 | 6 | 19 | 107 | 0% |
| DeepSeek R1 0528 | Frontier | 3 | 0 | 7 | 134 | 100% |
| Kimi K2.7 Code | Frontier | 0 | 3 | 1 | 8 | 0% |
| Claude Opus 4.8 | Frontier | 0 | 2 | 1 | 9 | 0% |
| DeepSeek V3.2 | Mid | 0 | 2 | 22 | 104 | 0% |
| GLM 5.2 | Frontier | 0 | 2 | 0 | 10 | 0% |
| MiMo V2.5 Pro | Frontier | 0 | 2 | 0 | 10 | 0% |
| Qwen3 Coder Next | Mid | 1 | 0 | 3 | 137 | 100% |
| DeepSeek V4 Flash | Mid | 0 | 1 | 1 | 10 | 0% |
| GPT 5.4 | Frontier | 0 | 1 | 0 | 131 | 0% |
| Kimi K2.5 | Frontier | 0 | 1 | 3 | 115 | 0% |
| MiniMax M3 | Frontier | 0 | 1 | 1 | 10 | 0% |
| Claude Haiku 4.5 | Small | 0 | 0 | 1 | 136 | n/a |
| DeepSeek V4 Pro | Frontier | 0 | 0 | 1 | 11 | n/a |
| Gemini 3.5 Flash | Small | 0 | 0 | 1 | 11 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 144 | n/a |
| GPT 5.5 | Frontier | 0 | 0 | 0 | 12 | n/a |
| Prompt | Tier | Arize AI | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 18 | 37 | 5 | 370 | 33% |
| ai-support-agent-platform | Beginner | 19 | 34 | 66 | 312 | 36% |
| ai-revenue-ops-copilot | Advanced | 23 | 9 | 2 | 385 | 72% |
| ai-revenue-ops-copilot | Beginner | 15 | 15 | 10 | 390 | 50% |
| ai-support-agent-platform | Intermediate | 20 | 8 | 5 | 395 | 71% |
| ai-revenue-ops-copilot | Intermediate | 12 | 2 | 4 | 406 | 86% |
| ai-agent-application | Intermediate | 0 | 5 | 0 | 15 | 0% |
| ai-agent-application | Advanced | 0 | 3 | 0 | 15 | 0% |
| ai-agent-application | Beginner | 0 | 3 | 5 | 12 | 0% |
| ai-engineering-workflow | Intermediate | 1 | 0 | 0 | 16 | 100% |
| ai-engineering-workflow | Advanced | 0 | 0 | 0 | 18 | n/a |
| ai-engineering-workflow | Beginner | 0 | 0 | 8 | 12 | n/a |
