An AI-first search API

One HTTP request for structured, traceable, model-friendly results.

Read the search API docs

The painful part of wiring search into an AI app is parsing the messy variety of pages and result layouts. TutuSoo collapses that into one request: GET /api/v1/search returns clean structured JSON.

Results carry title / snippet / url / source, support llms.txt, and fall back to the web when the local index is thin — so your model focuses on understanding and synthesis, not data cleaning.

Key features

One-line integration

GET /api/v1/search?q=… — no crawlers / parsers / anti-bot to manage.

Structured + traceable

Uniform results array, each tagged with source (local/brave) for citation.

llms.txt aware

Prefers a site’s llms.txt for higher-quality retrieval on AI-friendly sites.

Clear limits

Per-key rate limiting with X-RateLimit-* headers — quota at a glance.

How to use

  1. 1
    Get an API key
    Request one, or try the public per-IP tier for free first.
  2. 2
    Make a request
    GET /api/v1/search?q=your+question with your Authorization header.
  3. 3
    Feed your model
    Pass results as context to the LLM and keep source for citation.

FAQ

What format is returned?
Uniform JSON: a results array where each item has title / url / snippet / source.
How is Chinese support?
title/body are jieba-tokenized for better Chinese recall.
Can results be traced?
Yes, each carries a source tag (local index / brave fallback).
Is there a free allowance?
Yes, the public per-IP tier is free to try.

Integrate the TutuSoo API now

Free tier · structured results from one request · llms.txt supported

Read the search API docs