Web data for agents · by Tiny Fish, Inc.

TinyFish

Name: TinyFish
Author: Tiny Fish, Inc.

An API for AI agents that covers web search with JSON results, page fetching into Markdown/JSON/HTML, hosted Playwright/Puppeteer/CDP browser sessions, and a goal-driven agent that clicks through sites and returns structured output, all under one API key.

It gives an agent search, clean page text and a remote browser without running your own scrapers, proxies or headless Chrome fleet, and search and fetch cost nothing within rate limits.

APIFree tierCommercial: conditions

Open tinyfish.ai ↗ agent.md

For your project

Official API: Official API · Free tier Search: 30 requests/min, 500/hour free; Fetch: 150 URLs/min, 1,000/day free; Agent: 2 concurrent runs; Browser: 5 concurrent sessions (raised by contract)
“All REST API requests require an API key passed in the `X-API-Key` header.” docs.tinyfish.ai/authentication
Key: Needed
“All requests require the `X-API-Key` header.” docs.tinyfish.ai/search-api
Formats: JSON, MARKDOWN, HTML
Data download: Not stated
Price: Free tier Search and Fetch free within rate limits; new accounts start with $8 in Wallet funds for Agent and Browser
“Is there a monthly plan? No. There are no recurring Wallet plans. You fund the Wallet and pay for the Agent and Browser usage you actually run.” tinyfish.ai/pricing
License: Tiny Fish Terms of Service Access to the hosted services for your internal business purposes; no reselling or sublicensing the services
Commercial use: Commercial use with conditions
“a limited, worldwide, non-exclusive, non-transferable, non-sublicensable, revocable right to access and use the Services solely for your internal business purposes” tinyfish.ai/terms
Credit: Not stated
Embedding: Embedding unclear
Using the website: Free tier

Read on TinyFish's own pages on Oct 24, 2026 · license or terms · API docs · pricing. Check the license page before you ship.

For your agent

REST: send X-API-Key (env TINYFISH_API_KEY); SDKs are Python `from tinyfish import TinyFish` and TypeScript `@tiny-fish/sdk`; search is GET https://api.search.tinyfish.ai?query=...; MCP at https://agent.tinyfish.ai/mcp uses OAuth 2.1, no key.
Use is limited to your own internal business purposes; do not resell or sublicense the API itself
Stay within 30 searches/min and 150 fetched URLs/min on the free tier; Agent and Browser draw from a prepaid Wallet
Terms let Tiny Fish use run data to train its own models; do not send confidential data you cannot share

llms.txt: docs.tinyfish.ai/llms.txt · MCP: https://agent.tinyfish.ai/mcp

Paste into your agent: Read https://agentgid.com/resources/tinyfish/agent.md before using TinyFish.

Quick answers

Does TinyFish have an API?

Yes. TinyFish has an official API documented at https://docs.tinyfish.ai/.

Does the TinyFish API need a key?

Yes, you need an API key. Limits: Search: 30 requests/min, 500/hour free; Fetch: 150 URLs/min, 1,000/day free; Agent: 2 concurrent runs; Browser: 5 concurrent sessions (raised by contract)

Is TinyFish free?

Yes, there is a free tier: Search and Fetch free within rate limits; new accounts start with $8 in Wallet funds for Agent and Browser

Can I use TinyFish in a commercial project?

Yes, with conditions. Its terms say: "a limited, worldwide, non-exclusive, non-transferable, non-sublicensable, revocable right to access and use the Services solely for your internal business purposes"

What you can build

A research agent that searches, fetches the top pages as Markdown and cites them, using only the free Search and Fetch tiers
A price or availability watcher on pages without an API, using Monitor checks with webhook alerts
A form-filling back-office bot that runs Agent goals against supplier portals and stores the structured results

Alternatives in Web data for agents

All 5 →

AgentKey Instead of signing up for and wiring a separate paid API for search, scraping, social and market data, an agent gets all of them through one install and one bill, with fallback to a second provider when one fails. APIOpen sourceFree tierLicense unclear Context.dev An agent or RAG pipeline gets clean page text from JavaScript-heavy or protected sites without you running headless browsers and proxy pools. APICommercial: conditions Notte You get remote browsers that agents can log into and keep logged into, without running your own browser fleet, proxy pool or secret store. APIFree tierCommercial OK BrowserAct You get a scraper that runs again on a schedule and is meant to adapt when the page layout changes, instead of hand-maintaining CSS or XPath selectors. APIFree tierCommercial: conditions

Something wrong or out of date? Tell us.