Coding Agent x Model Benchmark Benchmark results for coding agents (OpenCode, Codex, OpenHands) across multiple LLM models. Each agent/model branch contains the agent's output for the Recipe Manager Web App task.