| Metric | Value |
|---|---|
| OpenAI wins | 1796 |
| LangChain wins | 972 |
| Abstains (no tool) | 128 |
| Other tool chosen | 1092 |
| Decisive cases | 2768 |
| OpenAI win rate (unweighted) | 64.9% |
| 95% CI | 63.1% - 66.6% |
| OpenAI win rate (weighted) | 64.9% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | OpenAI | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 108 | 46 | 9 | 45 | 70% |
| DeepSeek Pro 4–4 0813 | Frontier | 65 | 3 | 9 | 4 | 96% |
| DeepSeek R1 | Frontier | 174 | 11 | 23 | 2 | 94% |
| Devstral | Mid | 19 | 67 | 16 | 83 | 22% |
| Gemini Flash 2.5–3.8 | Small | 86 | 56 | 1 | 68 | 61% |
| Gemini Pro 2.5–3.1 Preview | Frontier | 91 | 116 | 2 | 3 | 44% |
| GLM 5 Turbo–5.3 | Frontier | 61 | 39 | 3 | 108 | 61% |
| GPT 5.4–6 Astra | Frontier | 186 | 0 | 0 | 21 | 100% |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 204 | 0 | 0 | 8 | 100% |
| GPT Mini 5.4 Mini–6 Luna | Mid | 191 | 0 | 1 | 19 | 100% |
| Haiku | Small | 18 | 4 | 0 | 190 | 82% |
| Kimi 2.5–3 | Frontier | 67 | 28 | 3 | 99 | 71% |
| Llama Maverick | Frontier | 0 | 161 | 27 | 18 | 0% |
| Llama Scout | Small | 0 | 129 | 11 | 63 | 0% |
| MiMo Pro 2–2.6 | Frontier | 158 | 25 | 4 | 23 | 86% |
| MiniMax M 2.7–3 | Frontier | 63 | 5 | 0 | 135 | 93% |
| Mistral Small | Mid | 12 | 61 | 18 | 83 | 16% |
| Opus 4.6–5.5 | Frontier | 125 | 66 | 0 | 21 | 65% |
| Qwen Coder | Mid | 93 | 99 | 1 | 17 | 48% |
| Sonnet 4.6–5.5 | Frontier | 75 | 56 | 0 | 82 | 57% |
| Prompt | Tier | OpenAI | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 259 | 188 | 0 | 144 | 58% |
| ai-support-agent-platform | Advanced | 258 | 186 | 1 | 144 | 58% |
| ai-revenue-ops-copilot | Intermediate | 196 | 214 | 1 | 177 | 48% |
| ai-revenue-ops-copilot | Beginner | 332 | 75 | 1 | 148 | 82% |
| ai-revenue-ops-copilot | Advanced | 185 | 221 | 4 | 163 | 46% |
| ai-support-agent-platform | Beginner | 308 | 55 | 1 | 203 | 85% |
| ai-agent-application | Advanced | 90 | 12 | 26 | 47 | 88% |
| ai-agent-application | Beginner | 89 | 13 | 50 | 26 | 87% |
| ai-agent-application | Intermediate | 79 | 8 | 44 | 40 | 91% |
