How it worksFeaturesDocsEarly accessGitHubDashboard

What Are Response Limits For?

Response Limits cap how large a tool's response can get before it's handed to the AI — maximum response length, JSON nesting depth, array item count, and error message length. They exist to stop one oversized upstream response from flooding an AI client's context window.

Open the Arthur screen that visually supports “What Are Response Limits For?”

Open the Settings tab to show the product area discussed by this wiki page

Learn more: How to Configure Response Limits.

Bring the Response Limits area into the viewport for a focused, semantic screenshot

Tutorial video