| Metric | Value |
|---|---|
| LangSmith wins | 1301 |
| Braintrust wins | 1456 |
| Abstains (no tool) | 216 |
| Other tool chosen | 1308 |
| Decisive cases | 2757 |
| LangSmith win rate (unweighted) | 47.2% |
| 95% CI | 45.3% - 49.1% |
| LangSmith win rate (weighted) | 47.2% |
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 | 108 | 17 | 26 | 72 | 86% |
| DeepSeek Pro 4–4 0813 | Frontier | 24 | 53 | 2 | 17 | 31% |
| DeepSeek R1 | Frontier | 190 | 0 | 11 | 27 | 100% |
| Devstral | Mid | 9 | 0 | 17 | 163 | 100% |
| Gemini Flash 2.5–3.8 | Small | 3 | 71 | 10 | 139 | 4% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 159 | 7 | 14 | 46 | 96% |
| GLM 5 Turbo–5.3 | Frontier | 22 | 140 | 19 | 47 | 14% |
| GPT 5.4–6 Astra | Frontier | 22 | 176 | 4 | 26 | 11% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 14 | 212 | 1 | 1 | 6% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 139 | 17 | 8 | 61 | 89% |
| Haiku | Small | 19 | 183 | 4 | 12 | 9% |
| Kimi 2.5–3 | Frontier | 1 | 179 | 5 | 29 | 1% |
| Llama Maverick | Frontier | 42 | 0 | 16 | 153 | 100% |
| Llama Scout | Small | 0 | 0 | 26 | 178 | n/a |
| MiMo Pro 2–2.6 | Frontier | 81 | 42 | 17 | 87 | 66% |
| MiniMax M 2.7–3 | Frontier | 64 | 109 | 12 | 40 | 37% |
| Mistral Small | Mid | 141 | 0 | 3 | 69 | 100% |
| Opus 4.6–5.5 | Frontier | 34 | 140 | 8 | 45 | 20% |
| Qwen Coder | Mid | 151 | 0 | 4 | 65 | 100% |
| Sonnet 4.6–5.5 | Frontier | 78 | 110 | 9 | 31 | 41% |
| Prompt | Tier | LangSmith | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Intermediate | 237 | 188 | 4 | 132 | 56% |
| ai-support-agent-platform | Intermediate | 270 | 137 | 6 | 148 | 66% |
| ai-revenue-ops-copilot | Advanced | 161 | 238 | 3 | 151 | 40% |
| ai-revenue-ops-copilot | Beginner | 160 | 224 | 12 | 167 | 42% |
| ai-support-agent-platform | Advanced | 181 | 196 | 5 | 184 | 48% |
| ai-support-agent-platform | Beginner | 125 | 153 | 77 | 210 | 45% |
| ai-agent-application | Advanced | 33 | 64 | 0 | 54 | 34% |
| ai-agent-application | Intermediate | 36 | 55 | 1 | 59 | 40% |
| ai-engineering-workflow | Intermediate | 19 | 72 | 1 | 58 | 21% |
| ai-engineering-workflow | Advanced | 19 | 70 | 1 | 59 | 21% |
| ai-agent-application | Beginner | 41 | 32 | 34 | 51 | 56% |
| ai-engineering-workflow | Beginner | 19 | 27 | 72 | 35 | 41% |
