AI Agent Evaluations
Performance results of AI coding agents on Next.js code generation and migration tasks, measuring success rate and execution time.
View on GitHub
Last run date: July 9, 2026
Agent Performance Results
Agent | ||||
|---|---|---|---|---|
Claude Fable 5 (high) | Claude Code | 233.93s | 92% | 96% |
Cursor Composer 2.5 | Cursor | 149.82s | 92% | 96% |
GPT 5.6 (xhigh) | Codex | 249.00s | 88% | 96% |
GPT 5.6 Sol (xhigh) | Codex | 164.45s | 88% | 96% |
GLM 5.2 | OpenCode | 285.17s | 88% | 96% |
Claude Opus 4.8 | Claude Code | 165.91s | 88% | 96% |
Grok 4.5 | OpenCode | 144.11s | 83% | 96% |
GPT 5.5 Pro | Codex | 771.63s | 83% | 83% |
GPT 5.4 (xhigh) | Codex | 220.42s | 83% | 88% |
GPT 5.3 Codex (xhigh) | Codex | 178.57s | 83% | 92% |
Kimi K2.7 Code | OpenCode | 207.72s | 75% | 96% |
MiniMax M3 | OpenCode | 181.88s | 75% | 96% |
GLM 5.1 | OpenCode | 253.24s | 75% | 96% |
Claude Opus 4.7 (max) | Claude Code | 142.48s | 75% | 96% |
Gemini 3.1 Pro Preview | Gemini CLI | 247.45s | 75% | 92% |
Claude Opus 4.6 | Claude Code | 189.27s | 75% | 96% |
Cursor Composer 2.0 | Cursor | 115.52s | 75% | 92% |
Kimi K2.6 | OpenCode | 197.69s | 67% | 96% |
Gemini 3.0 Pro Preview | Gemini CLI | 260.37s | 67% | 83% |
Cursor Composer 1.5 | Cursor | 119.20s | 67% | 88% |
Claude Sonnet 4.6 | Claude Code | 157.50s | 58% | 96% |
GPT 5.2 Codex (xhigh) | Codex | 149.60s | 58% | 79% |
MiniMax M2.7 | OpenCode | 293.14s | 50% | 63% |
Claude Sonnet 4.5 | Claude Code | 150.53s | 50% | 83% |
Kimi K2.5 | OpenCode | 134.99s | 21% | 58% |
* AGENTS.md provides bundled Next.js documentation for AI coding agents. The column shows additional evals that passed when agents had access to this documentation.
N/A marks evals a model has never run (newly added evals). They count against the success rate so all models are scored over the same eval set. Reworked evals keep a model's last result until it is rerun.