Search the hive
Firecrawl-compatible search. Blocked engines are retried through WARP and peer nodes, then merged.
Scrape a page
Turn any page into clean, LLM-ready markdown — same format as Firecrawl's /v2/scrape.
Map a site
List a website's URLs from its sitemaps, its links and a site search — fast, nothing is scraped.
Crawl a site
Scrape every page below a URL in the background, following links and sitemaps and respecting robots.txt.
AI extract & research
Turn pages into structured JSON, or let an agent search and read the web to answer a question. Uses this node's LLM only.
Settings
Configure this node. Changes apply immediately and are saved on this node only.
Node access (API key)
Decides who can use this node: the web UI, the Firecrawl SDKs and MCP all need it. Peers in the network never use it — they connect with the network key. Anyone with this key can run searches and use your AI credits, so share it only with your own devices and apps.
To change the key, set HC_API_KEYS in the node's .env and run docker compose up -d.
AI model (LLM)
This is a different key: your account with an AI provider. It stays on this node and is never shown again after saving. Used for Extract, Agent and the AI scrape formats — only for requests made with this node's API key, never for other nodes in the network. Works with Anthropic (Claude) and any OpenAI-compatible API.
Updates
- Version
- —
- Latest release
- —
- Last checked
- —
- Last updated
- —
- Auto-update
- —
Network
This node, its engines, and the peers it trades searches with. Refreshes every 5 seconds.
Engines
Peers
| Node | Route | Status | Answered | Asked / served | Priority | Agreement | Latency | Online |
|---|
Connect
Use this node from code, Firecrawl SDKs, or AI agents over MCP.