| Metric | Value |
|---|---|
| LangSmith wins | 1200 |
| Braintrust wins | 1243 |
| Abstains (no tool) | 183 |
| Other tool chosen | 1196 |
| Decisive cases | 2443 |
| LangSmith win rate (unweighted) | 49.1% |
| 95% CI | 47.1% - 51.1% |
| LangSmith win rate (weighted) | 49.1% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | LangSmith | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 10 | 193 | 0 | 1 | 5% |
| Claude Haiku 4.5 | Small | 18 | 162 | 3 | 11 | 10% |
| Claude Sonnet 4.6 | Frontier | 76 | 94 | 6 | 28 | 45% |
| DeepSeek R1 0528 | Frontier | 169 | 0 | 9 | 26 | 100% |
| Gemini 2.5 Pro | Frontier | 152 | 1 | 11 | 38 | 99% |
| GPT 5.4 Mini | Mid | 131 | 5 | 8 | 57 | 96% |
| Qwen3 Coder Next | Mid | 135 | 0 | 4 | 59 | 100% |
| Mistral Small 4 | Mid | 129 | 0 | 3 | 57 | 100% |
| Claude Opus 4.6 | Frontier | 1 | 113 | 0 | 18 | 1% |
| GPT 5.4 | Frontier | 18 | 92 | 0 | 22 | 16% |
| Kimi K2.5 | Frontier | 0 | 109 | 3 | 7 | 0% |
| GLM 5 Turbo | Frontier | 18 | 84 | 19 | 11 | 18% |
| MiniMax M2.7 | Frontier | 62 | 31 | 5 | 31 | 67% |
| DeepSeek V3.2 | Mid | 80 | 0 | 22 | 26 | 100% |
| GPT 5.5 | Frontier | 1 | 66 | 2 | 3 | 1% |
| MiMo V2 Pro | Frontier | 62 | 1 | 8 | 61 | 98% |
| MiniMax M3 | Frontier | 2 | 58 | 5 | 7 | 3% |
| DeepSeek V4 Pro | Frontier | 17 | 38 | 2 | 15 | 31% |
| Gemini 3.5 Flash | Small | 0 | 53 | 7 | 12 | 0% |
| Kimi K2.7 Code | Frontier | 0 | 53 | 1 | 17 | 0% |
| Claude Opus 4.8 | Frontier | 27 | 18 | 6 | 20 | 60% |
| GLM 5.2 | Frontier | 1 | 44 | 0 | 27 | 2% |
| MiMo V2.5 Pro | Frontier | 16 | 25 | 8 | 22 | 39% |
| Llama 4 Maverick | Frontier | 38 | 0 | 12 | 139 | 100% |
| DeepSeek V4 Flash | Mid | 25 | 3 | 4 | 39 | 89% |
| Devstral 2 2512 | Mid | 9 | 0 | 14 | 154 | 100% |
| Gemini 2.5 Flash | Small | 3 | 0 | 1 | 123 | 100% |
| Llama 4 Scout | Small | 0 | 0 | 20 | 165 | n/a |
| Prompt | Tier | LangSmith | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Intermediate | 225 | 167 | 4 | 126 | 57% |
| ai-support-agent-platform | Intermediate | 253 | 123 | 6 | 140 | 67% |
| ai-revenue-ops-copilot | Advanced | 152 | 216 | 3 | 143 | 41% |
| ai-revenue-ops-copilot | Beginner | 152 | 199 | 11 | 164 | 43% |
| ai-support-agent-platform | Advanced | 173 | 175 | 5 | 174 | 50% |
| ai-support-agent-platform | Beginner | 114 | 137 | 73 | 203 | 45% |
| ai-agent-application | Advanced | 25 | 45 | 0 | 42 | 36% |
| ai-agent-application | Intermediate | 29 | 38 | 1 | 45 | 43% |
| ai-engineering-workflow | Intermediate | 16 | 51 | 1 | 44 | 24% |
| ai-engineering-workflow | Advanced | 14 | 50 | 1 | 47 | 22% |
| ai-agent-application | Beginner | 33 | 22 | 25 | 39 | 60% |
| ai-engineering-workflow | Beginner | 14 | 20 | 53 | 29 | 41% |
