| Metric | Value |
|---|---|
| Braintrust wins | 806 |
| LangSmith wins | 902 |
| Abstains (no tool) | 105 |
| Other tool chosen | 862 |
| Decisive cases | 1708 |
| Braintrust win rate (unweighted) | 47.2% |
| 95% CI | 44.8% - 49.6% |
| Braintrust win rate (weighted) | 47.2% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Braintrust | LangSmith | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 136 | 8 | 0 | 0 | 94% |
| Claude Haiku 4.5 | Small | 113 | 14 | 1 | 9 | 89% |
| Claude Sonnet 4.6 | Frontier | 65 | 58 | 1 | 20 | 53% |
| DeepSeek R1 0528 | Frontier | 0 | 119 | 7 | 18 | 0% |
| Gemini 2.5 Pro | Frontier | 1 | 116 | 9 | 18 | 1% |
| Claude Opus 4.6 | Frontier | 113 | 1 | 0 | 18 | 99% |
| GPT 5.4 | Frontier | 92 | 18 | 0 | 22 | 84% |
| Kimi K2.5 | Frontier | 109 | 0 | 3 | 7 | 100% |
| GPT 5.4 Mini | Mid | 4 | 102 | 3 | 34 | 4% |
| GLM 5 Turbo | Frontier | 84 | 18 | 19 | 11 | 82% |
| Mistral Small 4 | Mid | 0 | 101 | 2 | 33 | 0% |
| Qwen3 Coder Next | Mid | 0 | 95 | 3 | 43 | 0% |
| MiniMax M2.7 | Frontier | 31 | 62 | 5 | 31 | 33% |
| DeepSeek V3.2 | Mid | 0 | 80 | 22 | 26 | 0% |
| MiMo V2 Pro | Frontier | 1 | 62 | 8 | 61 | 2% |
| Llama 4 Maverick | Frontier | 0 | 22 | 2 | 113 | 0% |
| GPT 5.5 | Frontier | 12 | 0 | 0 | 0 | 100% |
| MiniMax M3 | Frontier | 10 | 0 | 1 | 1 | 100% |
| Gemini 3.5 Flash | Small | 9 | 0 | 1 | 2 | 100% |
| GLM 5.2 | Frontier | 9 | 0 | 0 | 3 | 100% |
| DeepSeek V4 Pro | Frontier | 4 | 5 | 1 | 2 | 44% |
| MiMo V2.5 Pro | Frontier | 5 | 3 | 0 | 4 | 63% |
| Kimi K2.7 Code | Frontier | 6 | 0 | 1 | 5 | 100% |
| Claude Opus 4.8 | Frontier | 1 | 5 | 1 | 5 | 17% |
| Devstral 2 2512 | Mid | 0 | 6 | 6 | 123 | 0% |
| DeepSeek V4 Flash | Mid | 1 | 4 | 1 | 6 | 20% |
| Gemini 2.5 Flash | Small | 0 | 3 | 1 | 123 | 0% |
| Llama 4 Scout | Small | 0 | 0 | 7 | 124 | n/a |
| Prompt | Tier | Braintrust | LangSmith | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Intermediate | 130 | 189 | 4 | 101 | 41% |
| ai-support-agent-platform | Intermediate | 92 | 207 | 5 | 124 | 31% |
| ai-revenue-ops-copilot | Advanced | 164 | 128 | 2 | 125 | 56% |
| ai-revenue-ops-copilot | Beginner | 151 | 126 | 10 | 143 | 55% |
| ai-support-agent-platform | Advanced | 130 | 142 | 5 | 153 | 48% |
| ai-support-agent-platform | Beginner | 100 | 87 | 66 | 178 | 53% |
| ai-engineering-workflow | Advanced | 12 | 2 | 0 | 4 | 86% |
| ai-agent-application | Intermediate | 6 | 5 | 0 | 9 | 55% |
| ai-agent-application | Beginner | 3 | 8 | 5 | 4 | 27% |
| ai-engineering-workflow | Intermediate | 8 | 2 | 0 | 7 | 80% |
| ai-agent-application | Advanced | 7 | 3 | 0 | 8 | 70% |
| ai-engineering-workflow | Beginner | 3 | 3 | 8 | 6 | 50% |
