Mantle Firecrawl Agent

Your AI web scraping assistant

Connect Firecrawl and let the agent scrape pages, crawl sites, extract structured data, and research topics across the web — all in clean, LLM-ready format.

Connect your Firecrawl See all integrations

Free plan available · No credit card · API key connect

What the Firecrawl Agent does

Six things that make web data collection effortless. All backed by the Firecrawl engine.

Scrapes any page

Extract clean text, markdown, or structured data from any URL. Handles JavaScript rendering, anti-bot measures, and dynamic content.

Crawls entire sites

Point it at a domain and it recursively crawls every page, respecting your depth and URL filters.

Searches the web

Search for content across the web and get results back as clean, structured text ready for analysis.

Extracts structured data

Define a schema and Firecrawl extracts matching data from any set of pages at scale.

Runs deep research

Give it a topic and it researches across multiple sources, synthesizing findings into a comprehensive report.

Maps site structure

Discover all URLs on a domain and map the site architecture without crawling every page.

Three steps, two minutes

Get your API key. Connect it. Start extracting web data.

Step 1

Get your Firecrawl API key

Step 2

Connect to Mantle

Paste your API key in Settings and the agent connects to Firecrawl instantly.

Step 3

Start scraping

The agent can now scrape pages, crawl sites, and extract structured data on your behalf.

Powerful workflows, out of the box

Triggered by URLs, schedules, or research questions.

Scrape a competitor pageButton

One click to extract pricing, features, or content from any public web page.

Paste URLScrape pageExtract dataFormat results

Daily content monitorSchedule

Every morning, check a list of URLs for changes and report what is new.

Cron: daily 8amScrape target URLsCompare with previousPost changes

Research a topicButton

Give the agent a question and it scrapes, analyzes, and synthesizes an answer from multiple sources.

Define research questionSearch the webScrape top resultsCompile report

Map a websiteButton

One click to discover and list every URL on a domain. Great for SEO audits and content inventories.

Enter domainMap all URLsCategorize pagesExport sitemap

Every Firecrawl action, available as a tool

The agent can scrape, crawl, search, and extract — through the official Firecrawl API.

20tools

4categories

1API

Scrape

5 tools

Extract content from individual pages or batches of URLs.

Scrape page

Batch scrape

Get batch status

Cancel batch

Get batch errors

Crawl

5 tools

Recursively crawl websites with configurable depth and filters.

Start crawl

Get crawl status

Cancel crawl

List active crawls

Preview crawl params

Search & research

5 tools

Search the web and run deep research on any topic.

Web search

Deep research

Get research status

Start agent

Get agent status

Extract & map

5 tools

Extract structured data and map site architectures.

Extract data

Get extraction status

Map URLs

Generate LLMs.txt

Get credit usage

Your data pipeline, your rules

API key scoping and usage controls. Scrape responsibly.

API key authentication

All requests go through your Firecrawl account with your rate limits and usage quotas.

Usage monitoring

Track credit and token usage in real time. Set alerts before you hit limits.

Pause in one click

Toggle the agent off from the dashboard. All scraping stops immediately.

Revoke anytime

Rotate your API key from the Firecrawl dashboard to instantly cut all access.

Questions, answered

Firecrawl is a web scraping API that handles JavaScript rendering, anti-bot measures, and outputs clean markdown or structured data. It is built for LLM pipelines.

Yes. Firecrawl uses headless browsers to render dynamic content before extraction. SPAs, React apps, and lazy-loaded content all work.

Firecrawl respects robots.txt by default. You are responsible for complying with website terms of service and applicable laws.

Clean markdown by default. You can also get raw HTML, plain text, or structured JSON with a custom schema.

You can start for free — no credit card needed. The Free plan includes a usage allowance that refills every 5 hours. Paid plans unlock more usage, every AI model, and team features.