Firecrawl launches free index of 30 million US government pages for AI agents

What the index covers

The index spans federal, state, county and city government sources. Firecrawl lists several groups of material:

• Federal and state: statutes, regulations, court opinions, filings and agency guidance.
• County and city: municipal codes, zoning ordinances, permit requirements and fee schedules.
• Other public resources: tax forms, procurement notices and government publications.

The index is US-only for now. Firecrawl also says it connects pages with the laws and cases they cite, which is meant to help agents find related material instead of treating each page as an isolated result.

The company frames the problem as one of discovery, not availability. Most of this information is already public, it argues, but it sits across thousands of government websites, each with its own structure, search tools and collection of PDFs.

Why Firecrawl built it

Firecrawl says an AI agent asked a legal or regulatory question will often start with a web search, and the results may be a law firm's summary, a forum thread or a government page that points to another government page. Summaries can be out of date or refer to a different jurisdiction, the company notes, so being able to find the original source matters when the answer counts.

The company says teams building tax, compliance, permitting and legal research tools kept asking for a way to let their agents search official government information directly.

As an example, Firecrawl describes someone who wants to sell tacos from a window in Austin, Texas. Such a person might need state licensing requirements, fire inspection details and wastewater rules. According to the post, an agent can find the Texas health department's mobile food vendor guide, the Austin Fire Department's inspection information and Austin Water's wastewater rules, each on the agency's own website. For a location outside city limits, it can also surface the Travis County Fire Marshal's checklist and the relevant county code. Each result links back to its source.

Firecrawl is clear about limits: the index "won't replace a lawyer or the person behind the permit counter."

How to access it

Users can search the index from ChatGPT and Claude through the Firecrawl plugin. Firecrawl suggests questions such as which permits are needed to open a business in a given city, which regulations apply to a property, which court opinions interpret a statute, and whether there are local government contracts worth bidding on.

Developers have three routes: the /search/gov API endpoint, the firecrawl_gov_search tool in Firecrawl's hosted MCP server, and the firecrawl gov CLI command. Searches are free and do not use Firecrawl credits. No API key is needed to start, and adding one raises rate limits. Scraping a result to retrieve its full text uses standard scrape credits.

Firecrawl advises that queries work better when they include the relevant jurisdiction, and that naming an agency, citation or year can narrow results further.

What it means for users

The index joins the Developer Index and Research Index in Alexandria, Firecrawl's collection of searchable sources for AI agents. For people building research, tax, compliance or permitting tools, it offers a free way to point agents at primary government sources rather than secondary summaries.

Firecrawl cautions that, as with any search tool, results are not a guarantee that a rule is current or applies to a given situation. Users should check the source, jurisdiction and effective date before relying on an answer.