Leading: Vellum (57.1%)
| Metric | Value |
|---|---|
| Braintrust wins | 3 |
| Vellum wins | 4 |
| Abstains (no tool) | 22 |
| Other tool chosen | 980 |
| Decisive cases | 7 |
| Braintrust win rate (unweighted) | 42.9% |
| 95% CI | 15.8% - 75.0% |
| Braintrust win rate (weighted) | 42.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Braintrust | Vellum | None | Other | A rate |
|---|---|---|---|---|---|---|
| Devstral 2 2512 | Mid | 0 | 4 | 8 | 38 | 0% |
| Claude Haiku 4.5 | Small | 3 | 0 | 0 | 51 | 100% |
| Claude Opus 4.6 | Frontier | 0 | 0 | 0 | 54 | n/a |
| Claude Sonnet 4.6 | Frontier | 0 | 0 | 0 | 54 | n/a |
| DeepSeek R1 0528 | Frontier | 0 | 0 | 1 | 53 | n/a |
| DeepSeek V3.2 | Mid | 0 | 0 | 0 | 54 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 0 | 53 | n/a |
| Gemini 2.5 Pro | Frontier | 0 | 0 | 4 | 50 | n/a |
| GLM 5 Turbo | Frontier | 0 | 0 | 0 | 54 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 54 | n/a |
| GPT 5.4 | Frontier | 0 | 0 | 0 | 54 | n/a |
| GPT 5.4 Mini | Mid | 0 | 0 | 1 | 53 | n/a |
| Kimi K2.5 | Frontier | 0 | 0 | 1 | 47 | n/a |
| Llama 4 Maverick | Frontier | 0 | 0 | 0 | 54 | n/a |
| Llama 4 Scout | Small | 0 | 0 | 5 | 46 | n/a |
| MiMo V2 Pro | Frontier | 0 | 0 | 0 | 54 | n/a |
| MiniMax M2.7 | Frontier | 0 | 0 | 1 | 52 | n/a |
| Mistral Small 4 | Mid | 0 | 0 | 0 | 52 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 1 | 53 | n/a |
| Prompt | Tier | Braintrust | Vellum | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 0 | 4 | 1 | 166 | 0% |
| ai-revenue-ops-copilot | Beginner | 3 | 0 | 15 | 152 | 100% |
| ai-revenue-ops-copilot | Intermediate | 0 | 0 | 0 | 164 | n/a |
| ai-revenue-ops-copilot | Advanced | 0 | 0 | 1 | 163 | n/a |
| ai-support-agent-platform | Beginner | 0 | 0 | 5 | 164 | n/a |
| ai-support-agent-platform | Advanced | 0 | 0 | 0 | 171 | n/a |
