| Metric | Value |
|---|---|
| LangSmith wins | 1338 |
| Braintrust wins | 1577 |
| Abstains (no tool) | 237 |
| Other tool chosen | 1376 |
| Decisive cases | 2915 |
| LangSmith win rate (unweighted) | 45.9% |
| 95% CI | 44.1% - 47.7% |
| LangSmith win rate (weighted) | 45.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | LangSmith | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 109 | 29 | 26 | 72 | 79% |
| DeepSeek Pro 4–4 0813 | Frontier | 25 | 63 | 2 | 18 | 28% |
| DeepSeek R1 | Frontier | 200 | 0 | 12 | 28 | 100% |
| Devstral | Mid | 10 | 0 | 19 | 172 | 100% |
| Gemini Flash 2.5–3.8 | Small | 3 | 80 | 12 | 140 | 4% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 159 | 12 | 17 | 50 | 93% |
| GLM 5 Turbo–5.3 | Frontier | 23 | 147 | 19 | 51 | 14% |
| GPT 5.4–6 Astra | Frontier | 26 | 182 | 5 | 27 | 13% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 17 | 218 | 2 | 3 | 7% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 140 | 27 | 9 | 61 | 84% |
| Haiku | Small | 20 | 192 | 4 | 15 | 9% |
| Kimi 2.5–3 | Frontier | 1 | 189 | 5 | 31 | 1% |
| Llama Maverick | Frontier | 44 | 0 | 18 | 169 | 100% |
| Llama Scout | Small | 0 | 0 | 29 | 185 | n/a |
| MiMo Pro 2–2.6 | Frontier | 81 | 54 | 17 | 87 | 60% |
| MiniMax M 2.7–3 | Frontier | 64 | 119 | 12 | 41 | 35% |
| Mistral Small | Mid | 146 | 0 | 5 | 75 | 100% |
| Opus 4.6–5.5 | Frontier | 34 | 147 | 9 | 49 | 19% |
| Qwen Coder | Mid | 158 | 0 | 4 | 69 | 100% |
| Sonnet 4.6–5.5 | Frontier | 78 | 118 | 11 | 33 | 40% |
| Prompt | Tier | LangSmith | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Intermediate | 243 | 200 | 5 | 132 | 55% |
| ai-support-agent-platform | Intermediate | 276 | 146 | 6 | 152 | 65% |
| ai-revenue-ops-copilot | Advanced | 165 | 251 | 3 | 155 | 40% |
| ai-revenue-ops-copilot | Beginner | 162 | 237 | 13 | 172 | 41% |
| ai-support-agent-platform | Advanced | 184 | 209 | 5 | 189 | 47% |
| ai-support-agent-platform | Beginner | 126 | 163 | 80 | 215 | 44% |
| ai-agent-application | Advanced | 35 | 77 | 0 | 61 | 31% |
| ai-engineering-workflow | Intermediate | 22 | 84 | 1 | 66 | 21% |
| ai-agent-application | Intermediate | 39 | 64 | 1 | 71 | 38% |
| ai-engineering-workflow | Advanced | 21 | 80 | 1 | 66 | 21% |
| ai-agent-application | Beginner | 46 | 35 | 40 | 57 | 57% |
| ai-engineering-workflow | Beginner | 19 | 31 | 82 | 40 | 38% |
