Sprouts Data

npx add-mcp 'https://dashboard.sproutsdata.ai/mcp'

The Sprouts Data Platform (DSP) API answers questions about the business world: which companies exist in a market, who works there, and what they've patented, raised, published and are hiring for. Ask in plain English or with structured filters, and get back rows you can build on.

It searches Sprouts' business data: 43M+ companies, 200M+ people, and seven document collections — patents, news, trends, market reports, funding rounds, research papers and job openings.

[

Quickstart

Sign up, create a key and run your first search in five minutes.

](https://dashboard.sproutsdata.ai/docs/quickstart)[

API reference

Every endpoint, parameter and response field, with examples.

](https://dashboard.sproutsdata.ai/docs/api-reference/introduction)[

MCP server

The whole platform as tools inside Claude, Cursor or your own agent.

](https://dashboard.sproutsdata.ai/docs/mcp)[

For agents

The entire API on one page, written for LLM coding agents.

](https://dashboard.sproutsdata.ai/docs/for-agents)

What you can build

[

Prospect lists

Pageable lists of companies or people that match a definition you control.

](https://dashboard.sproutsdata.ai/docs/guides/search)[

Look-alike accounts

Net-new companies that resemble your best customers, ranked by similarity.

](https://dashboard.sproutsdata.ai/docs/guides/similar)[

CRM enrichment

Full records by domain or LinkedIn URL, with per-field provenance and honest misses.

](https://dashboard.sproutsdata.ai/docs/guides/entities)[

Market monitoring

Scheduled searches over funding rounds, news, patents and job openings.

](https://dashboard.sproutsdata.ai/docs/guides/search)[

Warm paths

The people and institutions that connect two companies, each with an honest confidence.

](https://dashboard.sproutsdata.ai/docs/guides/relationships)[

Account sizing

One company's people counted by department, seniority and location.

](https://dashboard.sproutsdata.ai/docs/guides/org-chart)

Honest coverage is the product

Every response tells you what was found, and also what the API could and could not do to find it:

• resolved_search shows how your query was interpreted.
• degradation[] names anything that limited the answer.
• total_is_estimate says whether the count is exact.

Explore the platform

[

Search

POST /v1/search — natural-language queries, structured filters and nine categories. Company descriptions are ranked by meaning; people are matched on filters and the parsed query.

](https://dashboard.sproutsdata.ai/docs/guides/search)[

Similar Companies

POST /v1/similar — look-alike companies from one or more anchors.

](https://dashboard.sproutsdata.ai/docs/guides/similar)[

Record Lookup

POST /v1/entities/* — full records by id, one route per category.

](https://dashboard.sproutsdata.ai/docs/guides/entities)[

Relationships

POST /v1/relationships* — who you already know somewhere, asked in structure or in words.

](https://dashboard.sproutsdata.ai/docs/guides/relationships)[

Sorting

sort on POST /v1/search orders results by up to three fields. GET /v1/sorts/{category} lists the fields, free to call.

](https://dashboard.sproutsdata.ai/docs/guides/search/sorting)[

Credits & usage

What each call costs, how a search is charged across pages, balance and per-key spend.

](https://dashboard.sproutsdata.ai/docs/guides/credits-and-usage)[

Usage history

GET /v1/usage/history — every run, with a free replayable snapshot of each search.

](https://dashboard.sproutsdata.ai/docs/guides/credits-and-usage/usage-history)[

Errors

RFC 9457 problem details with a machine-readable code from a closed set.

](https://dashboard.sproutsdata.ai/docs/guides/errors)

Base URL and auth

https://api.sproutsdata.ai

Authenticate every request with an API key in the X-API-Key header. Keys are created in the dashboard and are server-side only — never ship one in browser code.

curl -X POST https://api.sproutsdata.ai/v1/search \
  -H "X-API-Key: $SPROUTS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "fintech companies in berlin", "category": "company", "limit": 10}'

The nine categories

Category	What a result row is
company	A company in the entity index (an entity record)
person	A person in the entity index (an entity record)
patent	A patent document
news_article	A news article
trend	A trend report
market_report	A market report
funding_round	A funding-round record
academic_report	A research paper
job_opening	A job opening

The category you search is the category of every row you get back, so there is no mapping to maintain. Which categories are wired in a given environment is a runtime question: probe GET /v1/coverage rather than hard-coding the list.