| Metric | Value |
|---|---|
| Express wins | 22 |
| Next.js wins | 73 |
| Abstains (no tool) | 4 |
| Other tool chosen | 696 |
| Decisive cases | 95 |
| Express win rate (unweighted) | 23.2% |
| 95% CI | 15.8% - 32.6% |
| Express win rate (weighted) | 23.2% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Express | Next.js | None | Other | A rate |
|---|---|---|---|---|---|---|
| MiniMax M3 | Frontier | 0 | 18 | 0 | 24 | 0% |
| Mistral Small 4 | Mid | 0 | 11 | 0 | 24 | 0% |
| Llama 4 Scout | Small | 8 | 0 | 0 | 34 | 100% |
| Kimi K2.7 Code | Frontier | 0 | 8 | 0 | 33 | 0% |
| GLM 5.2 | Frontier | 0 | 7 | 0 | 35 | 0% |
| GPT 5.4 Mini | Mid | 0 | 7 | 0 | 34 | 0% |
| Llama 4 Maverick | Frontier | 6 | 0 | 1 | 28 | 100% |
| Qwen3 Coder Next | Mid | 1 | 5 | 0 | 36 | 17% |
| Devstral 2 2512 | Mid | 0 | 4 | 0 | 26 | 0% |
| Gemini 2.5 Pro | Frontier | 0 | 4 | 1 | 37 | 0% |
| GPT 5.5 | Frontier | 0 | 4 | 0 | 38 | 0% |
| DeepSeek R1 0528 | Frontier | 2 | 0 | 1 | 38 | 100% |
| MiMo V2.5 Pro | Frontier | 2 | 0 | 0 | 40 | 100% |
| DeepSeek V4 Flash | Mid | 1 | 1 | 0 | 32 | 50% |
| DeepSeek V4 Pro | Frontier | 1 | 1 | 0 | 39 | 50% |
| Gemini 3.5 Flash | Small | 0 | 2 | 1 | 39 | 0% |
| Claude Haiku 4.5 | Small | 1 | 0 | 0 | 34 | 100% |
| GPT 5.3 Codex | Frontier | 0 | 1 | 0 | 41 | 0% |
| Claude Opus 4.8 | Frontier | 0 | 0 | 0 | 42 | n/a |
| Claude Sonnet 4.6 | Frontier | 0 | 0 | 0 | 42 | n/a |
| Prompt | Tier | Express | Next.js | None | Other | A rate |
|---|---|---|---|---|---|---|
| full-stack-web-app | Beginner | 10 | 43 | 0 | 85 | 19% |
| full-stack-web-app | Intermediate | 10 | 23 | 0 | 103 | 30% |
| full-stack-web-app | Advanced | 2 | 6 | 1 | 98 | 25% |
| ai-agent-application | Beginner | 0 | 1 | 0 | 137 | 0% |
| ai-agent-application | Advanced | 0 | 0 | 2 | 136 | n/a |
| ai-agent-application | Intermediate | 0 | 0 | 1 | 137 | n/a |
