HushSearch documentation
One API for every kind of web data — search, scrape, crawl, browsers, documents and research.
HushSearch turns the web into data your code and agents can use: live search results, clean Markdown from any page, whole-site crawls, real browser sessions, PDF and OCR conversion, and cited research. Every endpoint takes the same X-API-Key header and reports what it cost and how long it took.
Quickstart
Get a key and make your first call in two minutes.
Authentication
API keys, the X-API-Key header and dashboard tokens.
Credits
How calls are priced, refunded and reported.
Playground
Try any endpoint in the browser, no key needed for the demo.
Your first request
curl -X POST https://api.hushsearch.io/v1/search/ \
-H "X-API-Key: $HUSHSEARCH_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query": "best open-source vector databases", "num_results": 5}'Every response includes credits_used, credits_remaining and response_time_ms.
Call paths exactly as documented, including the trailing slash (for example /v1/search/). The bare path answers with a redirect, which curl -X POST and browsers' CORS preflights do not follow.
For AI agents
Paste the prompt into Claude Code, Cursor or any coding agent, or add HushSearch as an MCP server. The complete reference for language models is at /llms.txt.
Add HushSearch to your agent
Links use a placeholder key; replace hs_live_YOUR_KEY after installing.
claude mcp add --transport http hushsearch https://api.hushsearch.io/v1/mcp/ --header 'X-API-Key: hs_live_YOUR_KEY'
Other MCP hosts (JSON config)
{
"mcpServers": {
"hushsearch": {
"url": "https://api.hushsearch.io/v1/mcp/",
"headers": {
"X-API-Key": "hs_live_YOUR_KEY"
}
}
}
}You have access to HushSearch, one API for all web data (search, scrape, crawl, browser, extract, research, documents).
Base URL: https://api.hushsearch.io
Authenticate every request with the header "X-API-Key: $HUSHSEARCH_API_KEY".
Every response includes credits_used, credits_remaining and response_time_ms.
Full endpoint reference for LLMs: http://localhost:3000/llms.txt
Start with POST /v1/search/ {"query": "..."}, POST /v1/scrape/ {"url": "..."} and POST /v1/answer/ {"query": "..."}.
Call paths exactly as documented, including the trailing slash, so no redirect is needed.
Prefer POST /v1/jobs/ for crawls and research that may take longer than 30 seconds.
On HTTP 402 the key is out of credits; on 429 back off and retry after the Retry-After header.{
"mcpServers": {
"hushsearch": {
"url": "https://api.hushsearch.io/v1/mcp/",
"headers": {
"X-API-Key": "hs_live_YOUR_KEY"
}
}
}
}Endpoints
35 endpoints, one key.
Search
| Web Search | POST /v1/search/ | 1 credit |
| Streaming Search | POST /v1/search/stream | 1 credit |
| Batch Search | POST /v1/search/batch | 1 credit / query |
| Find Similar | POST /v1/search/similar | 1 credit |
Answers & research
| Cited Answer | POST /v1/answer/ | 4 credits |
| Deep Research | POST /v1/research/ | 8 credits |
| Search & Summarize | POST /v1/search-and-summarize/ | 5 credits |
| Context Pack | POST /v1/context/ | 2 credits |
| Agent | POST /v1/agent/ | 4 credits |
Scrape & extract
| Scrape | POST /v1/scrape/ | 2 credits |
| Extract | POST /v1/extract/ | 3 credits |
| Reader | GET /v1/reader/ | 2 credits |
| Contents | POST /v1/contents/ | 2 credits / URL |
| Structured Extract | POST /v1/structured-extract/ | 4 credits |
| Search + Extract | POST /v1/search-extract/ | 6 credits |
| Multi-URL Extract | POST /v1/multi-extract/ | 4 credits |
Crawl & map
| Map | POST /v1/map/ | 1 credit |
| Crawl | POST /v1/crawl/ | 3 credits |
| Crawl Pipeline | POST /v1/crawl/pipeline | 4 credits |
| Batch Scrape | POST /v1/batch-scrape/ | 2 credits / URL |
Browsers
| Screenshot | POST /v1/screenshot/ | 3 credits |
| Browser Sessions | POST /v1/browser/sessions | 3 credits |
| Browser Execute | POST /v1/browser-execute/ | 2 credits |
| Agent Scrape | POST /v1/agent-scrape/ | 5 credits |
Documents
| Render → Docling | POST /v1/render-extract/ | 6 credits |
| Docling Markdown | POST /v1/docling-markdown/ | 4 credits |
| Docling OCR | POST /v1/docling-ocr/ | 3 credits |
| Docling Structured | POST /v1/docling-structured/ | 6 credits |
Platform
| Embeddings | POST /v1/embeddings/ | 1 credit / input |
| Rerank | POST /v1/rerank/ | 1 credit |
| Async Jobs | POST /v1/jobs/ | 5 credits / job (research 8 · crawl 5 · batch 3 · multi-extract 4) |
| Change Monitor | POST /v1/change-monitor/ | 2 credits |
| Knowledge Lookup | GET /v1/knowledge | 1 credit |
| Enrich | POST /v1/enrich | 4 credits |
| MCP Server | POST /v1/mcp/ | billed per tool call |