| Metric | Value |
|---|---|
| LangGraph wins | 35 |
| LangChain wins | 23 |
| Abstains (no tool) | 0 |
| Other tool chosen | 1 |
| Decisive cases | 58 |
| LangGraph win rate (unweighted) | 60.3% |
| 95% CI | 47.5% - 71.9% |
| LangGraph win rate (weighted) | 60.3% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | LangGraph | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | Small | 3 | 0 | 0 | 0 | 100% |
| Claude Opus 4.8 | Frontier | 3 | 0 | 0 | 0 | 100% |
| Gemini 3.5 Flash | Small | 3 | 0 | 0 | 0 | 100% |
| GLM 5.2 | Frontier | 3 | 0 | 0 | 0 | 100% |
| GPT 5.3 Codex | Frontier | 3 | 0 | 0 | 0 | 100% |
| GPT 5.4 Mini | Mid | 3 | 0 | 0 | 0 | 100% |
| GPT 5.5 | Frontier | 3 | 0 | 0 | 0 | 100% |
| Kimi K2.7 Code | Frontier | 3 | 0 | 0 | 0 | 100% |
| Claude Sonnet 4.6 | Frontier | 2 | 1 | 0 | 0 | 67% |
| DeepSeek V4 Pro | Frontier | 2 | 1 | 0 | 0 | 67% |
| MiMo V2.5 Pro | Frontier | 2 | 1 | 0 | 0 | 67% |
| MiniMax M3 | Frontier | 2 | 1 | 0 | 0 | 67% |
| Mistral Small 4 | Mid | 2 | 1 | 0 | 0 | 67% |
| DeepSeek R1 0528 | Frontier | 0 | 3 | 0 | 0 | 0% |
| Devstral 2 2512 | Mid | 0 | 3 | 0 | 0 | 0% |
| Gemini 2.5 Pro | Frontier | 0 | 3 | 0 | 0 | 0% |
| Llama 4 Maverick | Frontier | 0 | 3 | 0 | 0 | 0% |
| Qwen3 Coder Next | Mid | 0 | 3 | 0 | 0 | 0% |
| DeepSeek V4 Flash | Mid | 1 | 1 | 0 | 0 | 50% |
| Llama 4 Scout | Small | 0 | 2 | 0 | 1 | 0% |
| Prompt | Tier | LangGraph | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-agent-application | Intermediate | 13 | 7 | 0 | 0 | 65% |
| ai-agent-application | Advanced | 13 | 6 | 0 | 0 | 68% |
| ai-agent-application | Beginner | 9 | 10 | 0 | 1 | 47% |
