| Metric | Value |
|---|---|
| LangChain wins | 249 |
| OpenAI wins | 366 |
| Abstains (no tool) | 4 |
| Other tool chosen | 269 |
| Decisive cases | 615 |
| LangChain win rate (unweighted) | 40.5% |
| 95% CI | 36.7% - 44.4% |
| LangChain win rate (weighted) | 40.5% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | LangChain | OpenAI | None | Other | A rate |
|---|---|---|---|---|---|---|
| Claude Opus 4.6 | Frontier | 24 | 24 | 0 | 0 | 50% |
| GPT 5.3 Codex | Frontier | 0 | 48 | 0 | 0 | 0% |
| Gemini 2.5 Pro | Frontier | 33 | 14 | 1 | 0 | 70% |
| MiMo V2 Pro | Frontier | 8 | 39 | 0 | 0 | 17% |
| DeepSeek R1 0528 | Frontier | 3 | 42 | 0 | 2 | 7% |
| GPT 5.4 Mini | Mid | 0 | 45 | 0 | 2 | 0% |
| GPT 5.4 | Frontier | 0 | 43 | 0 | 5 | 0% |
| Qwen3 Coder Next | Mid | 17 | 25 | 0 | 6 | 40% |
| Llama 4 Maverick | Frontier | 41 | 0 | 0 | 6 | 100% |
| DeepSeek V3.2 | Mid | 13 | 22 | 0 | 12 | 37% |
| Gemini 2.5 Flash | Small | 17 | 17 | 0 | 14 | 50% |
| Claude Sonnet 4.6 | Frontier | 17 | 15 | 0 | 16 | 53% |
| Llama 4 Scout | Small | 24 | 0 | 0 | 23 | 100% |
| Devstral 2 2512 | Mid | 19 | 2 | 0 | 26 | 90% |
| Kimi K2.5 | Frontier | 8 | 11 | 1 | 23 | 42% |
| GLM 5 Turbo | Frontier | 14 | 2 | 2 | 30 | 88% |
| Mistral Small 4 | Mid | 10 | 3 | 0 | 25 | 77% |
| MiniMax M2.7 | Frontier | 0 | 10 | 0 | 36 | 0% |
| Claude Haiku 4.5 | Small | 1 | 4 | 0 | 43 | 20% |
| Prompt | Tier | LangChain | OpenAI | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Intermediate | 49 | 62 | 0 | 41 | 44% |
| ai-support-agent-platform | Advanced | 48 | 63 | 1 | 39 | 43% |
| ai-revenue-ops-copilot | Intermediate | 64 | 41 | 0 | 44 | 61% |
| ai-revenue-ops-copilot | Advanced | 63 | 42 | 2 | 40 | 60% |
| ai-revenue-ops-copilot | Beginner | 14 | 83 | 0 | 45 | 14% |
| ai-support-agent-platform | Beginner | 11 | 75 | 1 | 60 | 13% |
