# Agent Benchmark (https://app.infrasity.com/docs/dev-intelligence/agent-benchmark)





<AlphaNotice limitations="Coming soon. Agent Benchmark is still being built, and its scoring and layout may change before general release." />

Agent Benchmark tests how accurately models like Claude, GPT, and Gemini can generate and execute integration code directly from your documentation, then scores each model on whether the result passes.

<RelatedPages
  pages="[
{ title: &#x22;Dev Intelligence Overview&#x22;, description: &#x22;Track how developer-focused content performs in AI agent responses.&#x22;, href: &#x22;/dev-intelligence/overview&#x22; },
{ title: &#x22;Citation Check&#x22;, description: &#x22;Verify whether your content is cited by AI agents and LLMs.&#x22;, href: &#x22;/dev-intelligence/citation-check&#x22; },
{ title: &#x22;Agent Analytics&#x22;, description: &#x22;Analyze how AI agents interact with and cite your developer content.&#x22;, href: &#x22;/dev-intelligence/agent-analytics&#x22; },
]"
/>
