> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agentkey.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Company data for AI agents

> Crunchbase organizations through AgentKey: lookup, search, autocomplete, categories, locations, products, press references, awards and ownerships.

Company research starts with a reliable record — who a company is, where it
sits, what it makes, who owns it — and Crunchbase is the standard source,
normally behind an enterprise license. Through AgentKey your agent reads
Crunchbase organizations the same way it reads everything else —
`find_tools`, `describe_tool`, `execute_tool` — resolving a name to an
identifier and then pulling the record and whichever related cards it needs.

## What you can ask

* *"Find the Crunchbase record for Anthropic: description, headquarters and founders."*
* *"Which companies in the generative AI category are headquartered in London?"*
* *"List this company's subsidiaries and its parent."*
* *"Which organizations are most similar to Stripe?"*
* *"What products has this company launched, and when?"*
* *"Show press coverage of this company from the past month."*
* *"Which fintech companies won awards this year?"*
* *"这家公司的总部地址和分支机构在哪里？"*

## What's covered

**Finding an organization**

* Autocomplete a name to matching entity identifiers, scoped by entity type
* Search organizations with complex filters expressed in the request body
* Principals — organizations and people under one entity type — lookup and
  search

**The organization record**

* Full organization lookup by identifier
* Single cards on an organization: addresses, founders, awards, child
  organizations and ownerships, event appearances, acquisitions made and
  received, and funding, acquisition and closure predictions, among others
* Ownership records with parent and child organization cards and press
  references
* Organization similarity — lookup and search for comparable companies

**Taxonomy and geography**

* Categories, category groups and micro-categories — lookup and search
* Locations and addresses — lookup and search; address cards for the event or
  organization at that address

**Products, press, awards and events**

* Products, product launches and product similarities — lookup and search
* Press references and partnership announcements — lookup and search
* Awards and diversity spotlights — lookup and search
* Events and event appearances, with event cards for address, speakers,
  sponsors, exhibitors, organizers, contestants and press references

**Keeping a mirror current**

* Deleted entities, across all collections or for one collection
* Growth insights and research insights — lookup and search

## How to use it

Pass the request in plain language and let the router pick the tool:

```text theme={null}
find_tools(q="Crunchbase profile of Anthropic with founders and headquarters")
find_tools(q="伦敦的生成式 AI 公司有哪些")
find_tools(prefix="business/crunchbase")  # browse every Crunchbase entity type
describe_tool("<returned Provider/Operation>")
execute_tool("<Provider/Operation>", { ...params })
```

Both English and Chinese phrasing work. Always `describe_tool` before
executing — search tools take a structured query body, card lookups take a
card name, and the per-call cost is shown there. Resolve the identifier with
autocomplete first.

> [!NOTE]
>
> **Building with code?** The same capability is available as fixed-contract
> [Scenario Methods](/api-reference/scenario/overview) with automatic provider
> failover:
> [`business.company.search`](/api-reference/scenario/methods/business-company-search),
> [`business.company.get`](/api-reference/scenario/methods/business-company-get).

## FAQ

> **Do I need a Crunchbase license or API key?**
>
> No. AgentKey holds the upstream access; your agent only needs your
> AgentKey subscription.

> **How do I express a complex search?**
>
> Search tools accept a query in the request body — field filters, ordering
> and the fields to return. `describe_tool` shows the schema. For a simple
> name lookup, autocomplete returns the identifier directly.

> **How fresh is the data?**
>
> Every call goes to Crunchbase at request time; nothing is served from a
> stale archive. Records carry their own dates, and the deleted-entities
> tools let a local copy drop records Crunchbase has removed.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.