| Metric | Value |
|---|---|
| Vercel AI Gateway wins | 1 |
| AWS wins | 1 |
| Abstains (no tool) | 8 |
| Other tool chosen | 980 |
| Decisive cases | 2 |
| Vercel AI Gateway win rate (unweighted) | 50.0% |
| 95% CI | 9.5% - 90.5% |
| Vercel AI Gateway win rate (weighted) | 50.0% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model range | Tier | Vercel AI Gateway | AWS | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek Flash 3.2–4.1 | Mid | 0 | 0 | 0 | 54 | n/a |
| DeepSeek Pro 4–4 0813 | Frontier | 0 | 0 | 0 | 54 | n/a |
| DeepSeek R1 | Frontier | 0 | 0 | 1 | 53 | n/a |
| Devstral | Mid | 0 | 0 | 0 | 40 | n/a |
| Gemini Flash 2.5–3.8 | Small | 0 | 0 | 0 | 54 | n/a |
| Gemini Pro 2.5–3.1 Preview | Frontier | 0 | 0 | 0 | 54 | n/a |
| GLM 5 Turbo–5.3 | Frontier | 0 | 0 | 0 | 54 | n/a |
| GPT 5.4–6 Astra | Frontier | 0 | 0 | 0 | 50 | n/a |
| GPT Coding 5.3 Codex–6.1 Sol | Frontier | 0 | 0 | 0 | 53 | n/a |
| GPT Mini 5.4 Mini–6 Luna | Mid | 0 | 0 | 1 | 53 | n/a |
| Haiku | Small | 0 | 0 | 0 | 41 | n/a |
| Kimi 2.5–3 | Frontier | 0 | 0 | 0 | 42 | n/a |
| Llama Maverick | Frontier | 0 | 0 | 1 | 53 | n/a |
| Llama Scout | Small | 1 | 1 | 0 | 28 | 50% |
| MiMo Pro 2–2.6 | Frontier | 0 | 0 | 0 | 53 | n/a |
| MiniMax M 2.7–3 | Frontier | 0 | 0 | 0 | 51 | n/a |
| Mistral Small | Mid | 0 | 0 | 5 | 38 | n/a |
| Opus 4.6–5.5 | Frontier | 0 | 0 | 0 | 54 | n/a |
| Qwen Coder | Mid | 0 | 0 | 0 | 47 | n/a |
| Sonnet 4.6–5.5 | Frontier | 0 | 0 | 0 | 54 | n/a |
| Prompt | Tier | Vercel AI Gateway | AWS | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-agent-application | Advanced | 1 | 1 | 0 | 163 | 50% |
| ai-agent-application | Beginner | 0 | 0 | 1 | 170 | n/a |
| ai-agent-application | Intermediate | 0 | 0 | 1 | 166 | n/a |
| ai-engineering-workflow | Advanced | 0 | 0 | 0 | 156 | n/a |
| ai-engineering-workflow | Beginner | 0 | 0 | 6 | 166 | n/a |
| ai-engineering-workflow | Intermediate | 0 | 0 | 0 | 159 | n/a |
