| Metric | Value |
|---|---|
| DeepEval wins | 154 |
| Ragas wins | 182 |
| Abstains (no tool) | 183 |
| Other tool chosen | 3303 |
| Decisive cases | 336 |
| DeepEval win rate (unweighted) | 45.8% |
| 95% CI | 40.6% - 51.2% |
| DeepEval win rate (weighted) | 45.8% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | DeepEval | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| Llama 4 Maverick | Frontier | 108 | 0 | 12 | 69 | 100% |
| DeepSeek V4 Flash | Mid | 25 | 6 | 4 | 36 | 81% |
| MiMo V2 Pro | Frontier | 0 | 24 | 8 | 100 | 0% |
| MiniMax M2.7 | Frontier | 0 | 24 | 5 | 100 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 18 | 0 | 114 | 0% |
| GPT 5.4 Mini | Mid | 0 | 15 | 8 | 178 | 0% |
| Mistral Small 4 | Mid | 0 | 15 | 3 | 171 | 0% |
| GLM 5.2 | Frontier | 0 | 13 | 0 | 59 | 0% |
| Claude Sonnet 4.6 | Frontier | 0 | 12 | 6 | 186 | 0% |
| DeepSeek V4 Pro | Frontier | 4 | 7 | 2 | 59 | 36% |
| MiMo V2.5 Pro | Frontier | 3 | 8 | 8 | 52 | 27% |
| GPT 5.4 | Frontier | 9 | 1 | 0 | 122 | 90% |
| Claude Opus 4.8 | Frontier | 0 | 8 | 6 | 57 | 0% |
| GLM 5 Turbo | Frontier | 0 | 6 | 19 | 107 | 0% |
| Kimi K2.5 | Frontier | 4 | 1 | 3 | 111 | 80% |
| Kimi K2.7 Code | Frontier | 0 | 5 | 1 | 65 | 0% |
| Devstral 2 2512 | Mid | 0 | 4 | 14 | 159 | 0% |
| Llama 4 Scout | Small | 0 | 4 | 20 | 161 | 0% |
| MiniMax M3 | Frontier | 0 | 4 | 5 | 63 | 0% |
| DeepSeek V3.2 | Mid | 1 | 2 | 22 | 103 | 33% |
| Gemini 2.5 Pro | Frontier | 0 | 2 | 11 | 189 | 0% |
| GPT 5.5 | Frontier | 0 | 2 | 2 | 68 | 0% |
| Claude Haiku 4.5 | Small | 0 | 1 | 3 | 190 | 0% |
| DeepSeek R1 0528 | Frontier | 0 | 0 | 9 | 195 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 1 | 126 | n/a |
| Gemini 3.5 Flash | Small | 0 | 0 | 7 | 65 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 204 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 4 | 194 | n/a |
| Prompt | Tier | DeepEval | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 38 | 38 | 5 | 446 | 50% |
| ai-support-agent-platform | Beginner | 26 | 38 | 73 | 390 | 41% |
| ai-support-agent-platform | Intermediate | 23 | 9 | 6 | 484 | 72% |
| ai-agent-application | Intermediate | 5 | 24 | 1 | 83 | 17% |
| ai-agent-application | Beginner | 1 | 27 | 25 | 66 | 4% |
| ai-agent-application | Advanced | 10 | 16 | 0 | 86 | 38% |
| ai-revenue-ops-copilot | Beginner | 9 | 17 | 11 | 489 | 35% |
| ai-revenue-ops-copilot | Advanced | 13 | 10 | 3 | 488 | 57% |
| ai-engineering-workflow | Advanced | 10 | 0 | 1 | 101 | 100% |
| ai-revenue-ops-copilot | Intermediate | 8 | 2 | 4 | 508 | 80% |
| ai-engineering-workflow | Intermediate | 7 | 1 | 1 | 103 | 88% |
| ai-engineering-workflow | Beginner | 4 | 0 | 53 | 59 | 100% |
