| Metric | Value |
|---|---|
| Weights & Biases wins | 39 |
| Ragas wins | 47 |
| Abstains (no tool) | 36 |
| Other tool chosen | 878 |
| Decisive cases | 86 |
| Weights & Biases win rate (unweighted) | 45.3% |
| 95% CI | 35.3% - 55.8% |
| Weights & Biases win rate (weighted) | 45.3% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Weights & Biases | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| Gemini 2.5 Flash | Small | 19 | 0 | 0 | 33 | 100% |
| MiMo V2 Pro | Frontier | 3 | 13 | 2 | 36 | 19% |
| Devstral 2 2512 | Mid | 9 | 1 | 1 | 40 | 90% |
| MiniMax M2.7 | Frontier | 0 | 10 | 1 | 41 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 8 | 0 | 46 | 0% |
| GPT 5.4 Mini | Mid | 1 | 5 | 1 | 47 | 17% |
| Llama 4 Scout | Small | 4 | 1 | 3 | 41 | 80% |
| Mistral Small 4 | Mid | 0 | 4 | 0 | 47 | 0% |
| DeepSeek R1 0528 | Frontier | 3 | 0 | 2 | 49 | 100% |
| Claude Sonnet 4.6 | Frontier | 0 | 1 | 0 | 53 | 0% |
| DeepSeek V3.2 | Mid | 0 | 1 | 9 | 42 | 0% |
| GLM 5 Turbo | Frontier | 0 | 1 | 7 | 46 | 0% |
| GPT 5.4 | Frontier | 0 | 1 | 0 | 53 | 0% |
| Kimi K2.5 | Frontier | 0 | 1 | 0 | 47 | 0% |
| Claude Haiku 4.5 | Small | 0 | 0 | 1 | 51 | n/a |
| Gemini 2.5 Pro | Frontier | 0 | 0 | 6 | 48 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 54 | n/a |
| Llama 4 Maverick | Frontier | 0 | 0 | 0 | 53 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 3 | 51 | n/a |
| Prompt | Tier | Weights & Biases | Ragas | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Advanced | 17 | 6 | 1 | 140 | 74% |
| ai-support-agent-platform | Beginner | 4 | 17 | 25 | 123 | 19% |
| ai-support-agent-platform | Advanced | 6 | 13 | 1 | 149 | 32% |
| ai-revenue-ops-copilot | Beginner | 6 | 6 | 4 | 152 | 50% |
| ai-support-agent-platform | Intermediate | 4 | 3 | 4 | 155 | 57% |
| ai-revenue-ops-copilot | Intermediate | 2 | 2 | 1 | 159 | 50% |
