| Metric | Value |
|---|---|
| DeepEval wins | 179 |
| Ragas wins | 215 |
| Abstains (no tool) | 237 |
| Other tool chosen | 3897 |
| Decisive cases | 394 |
| DeepEval win rate (unweighted) | 45.4% |
| 95% CI | 40.6% - 50.4% |
| DeepEval win rate (weighted) | 45.4% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | DeepEval | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 29 | 9 | 26 | 172 | 76% |
| DeepSeek Pro 4–4 0813 | Frontier | 5 | 7 | 2 | 94 | 42% |
| DeepSeek R1 | Frontier | 0 | 0 | 12 | 228 | n/a |
| Devstral | Mid | 5 | 4 | 19 | 173 | 56% |
| Gemini Flash 2.5–3.8 | Small | 0 | 2 | 12 | 221 | 0% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 0 | 5 | 17 | 216 | 0% |
| GLM 5 Turbo–5.3 | Frontier | 0 | 25 | 19 | 196 | 0% |
| GPT 5.4–6 Astra | Frontier | 9 | 3 | 5 | 223 | 75% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 0 | 0 | 2 | 238 | n/a |
| GPT Mini 5.4 Mini–6 Luna | Mid | 0 | 17 | 9 | 211 | 0% |
| Haiku | Small | 0 | 1 | 4 | 226 | 0% |
| Kimi 2.5–3 | Frontier | 4 | 10 | 5 | 207 | 29% |
| Llama Maverick | Frontier | 123 | 0 | 18 | 90 | 100% |
| Llama Scout | Small | 0 | 5 | 29 | 180 | 0% |
| MiMo Pro 2–2.6 | Frontier | 4 | 32 | 17 | 186 | 11% |
| MiniMax M 2.7–3 | Frontier | 0 | 31 | 12 | 193 | 0% |
| Mistral Small | Mid | 0 | 19 | 5 | 202 | 0% |
| Opus 4.6–5.5 | Frontier | 0 | 31 | 9 | 199 | 0% |
| Qwen Coder | Mid | 0 | 0 | 4 | 227 | n/a |
| Sonnet 4.6–5.5 | Frontier | 0 | 14 | 11 | 215 | 0% |
| Prompt | Tier | DeepEval | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 41 | 39 | 5 | 502 | 51% |
| ai-support-agent-platform | Beginner | 28 | 39 | 80 | 437 | 42% |
| ai-agent-application | Intermediate | 7 | 36 | 1 | 131 | 16% |
| ai-agent-application | Beginner | 1 | 41 | 40 | 96 | 2% |
| ai-support-agent-platform | Intermediate | 24 | 9 | 6 | 541 | 73% |
| ai-agent-application | Advanced | 14 | 19 | 0 | 140 | 42% |
| ai-revenue-ops-copilot | Beginner | 11 | 17 | 13 | 543 | 39% |
| ai-revenue-ops-copilot | Advanced | 13 | 11 | 3 | 547 | 54% |
| ai-engineering-workflow | Advanced | 15 | 0 | 1 | 152 | 100% |
| ai-engineering-workflow | Intermediate | 11 | 1 | 1 | 160 | 92% |
| ai-revenue-ops-copilot | Intermediate | 9 | 2 | 5 | 564 | 82% |
| ai-engineering-workflow | Beginner | 5 | 1 | 82 | 84 | 83% |
