TinyFish

The web operating layer for AI agents

Documentation

TinyFish Developer Documentation

Choose the right TinyFish API, authenticate quickly, and start integrating faster

TinyFish gives you four public API surfaces: Agent, Search, Fetch, and Browser. This docs site is organized to help you pick the right one quickly, authenticate once, and move straight to the examples or reference you need.

Free Search and Fetch. Credits are used only by Agent and Browser.

Start Here

Run your first integration in minutes Set up API keys and MCP auth Single-page context for Claude, Codex, and Cursor Debug auth, rate limits, and run failures

Choose the Right API

Use natural-language goals to automate workflows on real sites Get structured ranked web results for a query Fetch URLs and extract clean page content Create a remote browser session for direct Playwright/CDP control

Canonical Endpoints

ProductCanonical endpoint
Agenthttps://agent.tinyfish.ai/v1/automation/...
SearchGET https://api.search.tinyfish.ai
FetchPOST https://api.fetch.tinyfish.ai
BrowserPOST https://api.browser.tinyfish.ai

Learn the Basics

Pick `/run`, `/run-async`, or `/run-sse` Understand lifecycle, polling, and statuses Write instructions that produce reliable results Keep logged-in state across runs Choose lite vs stealth Route runs through supported countries Connect TinyFish to Claude and other assistants