Leading: Anthropic Claude API (62.8%)
| Metric | Value |
|---|---|
| AI SDK wins | 16 |
| Anthropic Claude API wins | 27 |
| Abstains (no tool) | 4 |
| Other tool chosen | 1164 |
| Decisive cases | 43 |
| AI SDK win rate (unweighted) | 37.2% |
| 95% CI | 24.4% - 52.1% |
| AI SDK win rate (weighted) | 37.2% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | AI SDK | Anthropic Claude API | None | Other | A rate |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | Small | 0 | 18 | 0 | 48 | 0% |
| MiniMax M2.7 | Frontier | 1 | 9 | 0 | 50 | 10% |
| GLM 5 Turbo | Frontier | 7 | 0 | 2 | 57 | 100% |
| Devstral 2 2512 | Mid | 6 | 0 | 0 | 57 | 100% |
| DeepSeek V3.2 | Mid | 2 | 0 | 0 | 63 | 100% |
| Claude Opus 4.6 | Frontier | 0 | 0 | 0 | 66 | n/a |
| Claude Sonnet 4.6 | Frontier | 0 | 0 | 0 | 66 | n/a |
| DeepSeek R1 0528 | Frontier | 0 | 0 | 0 | 64 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 0 | 66 | n/a |
| Gemini 2.5 Pro | Frontier | 0 | 0 | 1 | 65 | n/a |
| GPT 5.3 Codex | Frontier | 0 | 0 | 0 | 66 | n/a |
| GPT 5.4 | Frontier | 0 | 0 | 0 | 66 | n/a |
| GPT 5.4 Mini | Mid | 0 | 0 | 0 | 63 | n/a |
| Kimi K2.5 | Frontier | 0 | 0 | 1 | 57 | n/a |
| Llama 4 Maverick | Frontier | 0 | 0 | 0 | 65 | n/a |
| Llama 4 Scout | Small | 0 | 0 | 0 | 64 | n/a |
| MiMo V2 Pro | Frontier | 0 | 0 | 0 | 65 | n/a |
| Mistral Small 4 | Mid | 0 | 0 | 0 | 50 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 0 | 66 | n/a |
| Prompt | Tier | AI SDK | Anthropic Claude API | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Advanced | 1 | 14 | 1 | 190 | 7% |
| ai-revenue-ops-copilot | Advanced | 0 | 10 | 2 | 186 | 0% |
| ai-support-agent-platform | Beginner | 9 | 0 | 1 | 189 | 100% |
| ai-support-agent-platform | Intermediate | 6 | 1 | 0 | 201 | 86% |
| ai-revenue-ops-copilot | Intermediate | 0 | 1 | 0 | 204 | 0% |
| ai-revenue-ops-copilot | Beginner | 0 | 1 | 0 | 194 | 0% |