Next.js Agent Evals (Codex) - Success Rate with AGENTS.md: leaderboard
Vercel's Next.js Agent Evals run coding agents on 31 real Next.js tasks; success is pass@4, infrastructure failures are rerun and evals an agent never ran count as failures. This board is the percentage of evals passed by models in the Codex harness with Next.js documentation bundled in AGENTS.md; higher is better.
Metric: Evals passed with bundled Next.js docs in AGENTS.md, pass@4 (%). Source: nextjs.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra (High) | 100 | #6 (GPT-6 Astra) |
| 2 | GPT-6.1 Terra (High) | 100 | #15 (GPT-6.1 Terra) |
| 3 | GPT-6 Sol (High) | 100 | #19 (GPT-6 Sol) |
| 4 | GPT-6 Luna (High) | 96 | #61 (GPT-6 Luna) |
| 5 | GPT-5.3 Codex (xHigh) | 75 | #72 (GPT-5.3 Codex) |
| 6 | GPT-5.4 (xHigh) | 71 | #83 (GPT-5.4) |
| 7 | GPT-5.5 Pro | 68 | #23 |
| 8 | GPT-5.2 Codex (xHigh) | 64 | #99 (GPT-5.2 Codex) |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=next-js-agent-evals-codex-success-rate-with-agents-md · How It Works · Data refreshed daily, snapshot 2026-11-05.