| Metric | Value |
|---|---|
| OpenAI wins | 1571 |
| LangChain wins | 912 |
| Abstains (no tool) | 103 |
| Other tool chosen | 932 |
| Decisive cases | 2483 |
| OpenAI win rate (unweighted) | 63.3% |
| 95% CI | 61.4% - 65.1% |
| OpenAI win rate (weighted) | 63.3% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | OpenAI | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| Gemini 2.5 Pro | Frontier | 72 | 113 | 1 | 2 | 39% |
| GPT 5.3 Codex | Frontier | 183 | 0 | 0 | 5 | 100% |
| GPT 5.4 Mini | Mid | 175 | 0 | 1 | 11 | 100% |
| Qwen3 Coder Next | Mid | 84 | 87 | 1 | 14 | 49% |
| DeepSeek R1 0528 | Frontier | 157 | 10 | 18 | 1 | 94% |
| Llama 4 Maverick | Frontier | 0 | 148 | 21 | 15 | 0% |
| Claude Sonnet 4.6 | Frontier | 72 | 54 | 0 | 63 | 57% |
| MiMo V2 Pro | Frontier | 109 | 13 | 0 | 2 | 89% |
| Claude Opus 4.6 | Frontier | 60 | 62 | 0 | 4 | 49% |
| GPT 5.4 | Frontier | 117 | 0 | 0 | 9 | 100% |
| Llama 4 Scout | Small | 0 | 115 | 7 | 57 | 0% |
| Gemini 2.5 Flash | Small | 55 | 42 | 0 | 29 | 57% |
| DeepSeek V3.2 | Mid | 58 | 38 | 0 | 29 | 60% |
| Devstral 2 2512 | Mid | 11 | 65 | 12 | 74 | 14% |
| Mistral Small 4 | Mid | 11 | 56 | 14 | 71 | 16% |
| Claude Opus 4.8 | Frontier | 59 | 1 | 0 | 2 | 98% |
| Kimi K2.5 | Frontier | 26 | 27 | 1 | 59 | 49% |
| DeepSeek V4 Pro | Frontier | 49 | 3 | 9 | 2 | 94% |
| GPT 5.5 | Frontier | 49 | 0 | 0 | 8 | 100% |
| GLM 5.2 | Frontier | 41 | 5 | 1 | 16 | 89% |
| MiMo V2.5 Pro | Frontier | 33 | 12 | 4 | 13 | 73% |
| DeepSeek V4 Flash | Mid | 34 | 6 | 8 | 11 | 85% |
| GLM 5 Turbo | Frontier | 7 | 33 | 2 | 83 | 18% |
| MiniMax M3 | Frontier | 29 | 3 | 0 | 30 | 91% |
| Gemini 3.5 Flash | Small | 17 | 12 | 1 | 31 | 59% |
| MiniMax M2.7 | Frontier | 24 | 2 | 0 | 88 | 92% |
| Kimi K2.7 Code | Frontier | 22 | 1 | 2 | 36 | 96% |
| Claude Haiku 4.5 | Small | 17 | 4 | 0 | 167 | 81% |
| Prompt | Tier | OpenAI | LangChain | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 233 | 173 | 1 | 122 | 57% |
| ai-support-agent-platform | Intermediate | 224 | 178 | 0 | 129 | 56% |
| ai-revenue-ops-copilot | Beginner | 301 | 74 | 0 | 127 | 80% |
| ai-revenue-ops-copilot | Intermediate | 170 | 201 | 1 | 157 | 46% |
| ai-revenue-ops-copilot | Advanced | 160 | 208 | 3 | 143 | 43% |
| ai-support-agent-platform | Beginner | 274 | 52 | 1 | 181 | 84% |
| ai-agent-application | Advanced | 75 | 9 | 22 | 29 | 89% |
| ai-agent-application | Beginner | 71 | 11 | 39 | 17 | 87% |
| ai-agent-application | Intermediate | 63 | 6 | 36 | 27 | 91% |
