| Metric | Value |
|---|---|
| LangChain wins | 905 |
| OpenAI wins | 1525 |
| Abstains (no tool) | 93 |
| Other tool chosen | 938 |
| Decisive cases | 2430 |
| LangChain win rate (unweighted) | 37.2% |
| 95% CI | 35.3% - 39.2% |
| LangChain win rate (weighted) | 37.2% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | LangChain | OpenAI | None | Other | A rate |
|---|---|---|---|---|---|---|
| Gemini 2.5 Pro | Frontier | 111 | 70 | 2 | 2 | 61% |
| GPT 5.3 Codex | Frontier | 0 | 180 | 0 | 5 | 0% |
| GPT 5.4 Mini | Mid | 0 | 173 | 1 | 10 | 0% |
| Qwen3 Coder Next | Mid | 87 | 80 | 1 | 15 | 52% |
| DeepSeek R1 0528 | Frontier | 11 | 155 | 15 | 2 | 7% |
| Llama 4 Maverick | Frontier | 147 | 0 | 18 | 15 | 100% |
| Claude Opus 4.6 | Frontier | 65 | 63 | 0 | 4 | 51% |
| MiMo V2 Pro | Frontier | 13 | 115 | 0 | 2 | 10% |
| Claude Sonnet 4.6 | Frontier | 53 | 70 | 0 | 63 | 43% |
| GPT 5.4 | Frontier | 0 | 122 | 0 | 10 | 0% |
| Llama 4 Scout | Small | 110 | 0 | 6 | 61 | 100% |
| Gemini 2.5 Flash | Small | 44 | 57 | 0 | 31 | 44% |
| DeepSeek V3.2 | Mid | 40 | 60 | 0 | 31 | 40% |
| Devstral 2 2512 | Mid | 67 | 11 | 12 | 78 | 86% |
| Mistral Small 4 | Mid | 53 | 11 | 12 | 73 | 83% |
| Kimi K2.5 | Frontier | 27 | 28 | 1 | 62 | 49% |
| Claude Opus 4.8 | Frontier | 1 | 51 | 0 | 1 | 2% |
| DeepSeek V4 Pro | Frontier | 3 | 40 | 9 | 2 | 7% |
| GPT 5.5 | Frontier | 0 | 43 | 0 | 6 | 0% |
| GLM 5 Turbo | Frontier | 34 | 7 | 2 | 88 | 83% |
| MiMo V2.5 Pro | Frontier | 10 | 28 | 4 | 11 | 26% |
| GLM 5.2 | Frontier | 4 | 34 | 1 | 15 | 11% |
| DeepSeek V4 Flash | Mid | 5 | 29 | 7 | 9 | 15% |
| MiniMax M3 | Frontier | 3 | 24 | 0 | 26 | 11% |
| Gemini 3.5 Flash | Small | 11 | 15 | 1 | 25 | 42% |
| MiniMax M2.7 | Frontier | 2 | 24 | 0 | 94 | 8% |
| Kimi K2.7 Code | Frontier | 1 | 19 | 1 | 31 | 5% |
| Claude Haiku 4.5 | Small | 3 | 16 | 0 | 166 | 16% |
| Prompt | Tier | LangChain | OpenAI | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 174 | 230 | 1 | 124 | 43% |
| ai-support-agent-platform | Intermediate | 174 | 223 | 0 | 134 | 44% |
| ai-revenue-ops-copilot | Intermediate | 205 | 167 | 1 | 156 | 55% |
| ai-revenue-ops-copilot | Beginner | 71 | 300 | 0 | 129 | 19% |
| ai-revenue-ops-copilot | Advanced | 208 | 158 | 4 | 144 | 57% |
| ai-support-agent-platform | Beginner | 50 | 272 | 1 | 187 | 16% |
| ai-agent-application | Advanced | 9 | 63 | 18 | 26 | 13% |
| ai-agent-application | Beginner | 9 | 60 | 35 | 15 | 13% |
| ai-agent-application | Intermediate | 5 | 52 | 33 | 23 | 9% |
