Indexed API

1. 1

   Create a key

   Open API keys, create a key and save it when it appears. Prefer a browser test? Click Test key there to see a response and its actual credit cost.
2. 2

   Set your key in your terminal

   The examples read the key from an environment variable. Run this in your terminal, replacing the placeholder with your key. Keep it on your server or local machine; exclude it from browser code and public repositories.

export INDEXED_API_KEY='idx_YOUR_KEY_HERE'

Look up stripe.com, or replace it with a company domain you want to research. This name/domain search works on every plan. Free includes 25 searches a day; within your quota this request costs zero credits. Filter-only searches require a paid plan.

curl --fail-with-body --max-time 30 'https://indexed.vc/api/v1/companies?domain=stripe.com&limit=1' \
  -H "X-API-Key: $INDEXED_API_KEY"

curl prints the JSON response. With Node.js 18 or later, save the example as lookup.mjs or the Python example as lookup.py and run the command shown; neither needs an additional library.

Read the result

A match returns a public company card in data. _revealed: false means funding-history and other depth remain locked; the public card is still usable. Detail/reveal endpoints unlock depth using credits. This is an Illustrative response excerpt; returned fields and current values vary by company.

{
  "data": [
    {
      "name": "Stripe",
      "slug": "stripe",
      "website": "https://stripe.com",
      "_revealed": false
    }
  ]
}

A successful quota-included search returns X-Credits-Cost: 0. Your actual quota and request limits are in the response headers. Add -i to curl to see them, and retain X-Request-Id when investigating a failure.

Next: get company depth, look up a list of domains, or check usage.

Indexed focuses on technology and product companies globally, including bootstrapped businesses, and their funding and investor relationships. Software, hardware, platforms and scientific or engineering products are core areas. Coverage and field completeness vary by company and market; a missing result is not an eligibility decision.

Agencies, consultancies and ordinary nontechnology businesses are outside our current focus. Funding, company size or using software does not by itself establish fit. Mixed services and product businesses require evidence of what the exact company owns and sells; ambiguous cases are held for review.

Additions require verified identity and supporting evidence. Existing coverage is not a complete registry or a fully scope-reviewed catalogue. Decisions can be reviewed again when new product or identity evidence becomes available.

An unfiltered, first-page lookup using the explicit domain parameter returns an empty data array on a miss and a top-level coverage_requested boolean. true means the request was durably recorded for manual review; false means recording failed and you can retry. Neither admission nor a review date is guaranteed. Batch lookup reports this boolean on each miss item, with coverage_eta_days: null. Filtered, paginated and q-based searches do not return this acknowledgement; its absence does not mean recording failed.

For a company-domain miss, check meta.alternate_entities when present: the domain may belong to a visible investor with its own profile. Use the returned profile_path to identify that profile.

The Indexed API uses API keys for authentication. Pass your key in the X-API-Key header on every request. All requests must be made over HTTPS. Calls without authentication will fail.

API keys start with idx_. Keep keys secret: do not expose them in browser-side code, public repositories, or client bundles. If a key is compromised, revoke it immediately from your API keys dashboard.

X-API-Key: idx_YOUR_KEY_HERE

Enriching a list of company domains? The domain to company guide covers single and batch lookups, matching rules and what misses mean. For prices and coverage next to Crunchbase, PitchBook and others, see the comparison.

Search is included. Every account has a daily search quota that costs no credits: free: 25; silver: 250; gold: 1,000; platinum: 2,500; enterprise: 10,000. Enterprise values are defaults; contractual overrides may apply. The quota is per account across all of its keys and resets at 00:00 UTC. Past the quota a paid plan continues at 1 credit per page; the Free plan waits for the reset. A search that returns nothing never counts and is never charged. Every response carries meta.search_quota and X-Search-Quota-Remaining.

Depth — funding history, valuations, stacks, investor relationships — costs credits from your monthly allocation on every plan, Free included. All tiers use one unified credit pool, the same credits that power profile views, exports, and tech stack reveals in the dashboard; an entity you reveal in the app is _revealed in the API and vice versa. On Free,25 credits a month is about 8 full records, refreshed monthly.

Entity cache: repeating a detail fetch for the same entity at the same depth is free for 12 months. A different depth or endpoint is not automatically free. Reveal and enrich share the existing entity reveal cache; search's _revealedflag reports that state, not a promise that every detail action is free. Concurrent requests honor these same cache rules without charging again for a cached action. Internal batch recovery reuses its original charge; a new HTTP request is a separate operation subject to these cache rules. While a search page still has unrevealed rows, its meta.depth names the detail endpoint that unlocks them and what it costs, so a client can act on a public card without reading these docs.

Out of credits? The response is a 402 CREDIT_LIMIT_REACHED carryingtop_up_url, upgrade_url and next_plan (the plan one step up, with its credits and price), so a client can recover without leaving the API. Enable overage at $0.065/credit in Settings, or buy a credit pack (500 credits for $25, up to 25% off at volume).

On every plan a search result is the full public card — identical to what an anonymous visitor sees on indexed.vc, funding summary included. What you pay to unlock is depth: funding history, valuations, tech/app stacks, and investor relationships, via POST /reveal or the entity detail endpoints — on the Free plan too, from its monthly credits. Repeat access follows the action-specific cache rules above.

Each result in the data array carries a reveal flag:

The meta block on search responses includes revealed_count and unrevealed_count for the current page, so you know how many new reveals you'd need to unlock every entity's depth.

Exact company-domain misses may also include meta.alternate_entities when the domain belongs to a visible investor. These are identity-only navigation pointers; they never expose investor AUM, thesis, check sizes, contacts, portfolio, or deal data, and they do not turn an investor into a company match.

Filter by reveal state with the reveal_status query param:

Every value is an ordinary search: free inside your daily quota, 1 credit per nonempty page past it on paid plans. An empty data page consumes no search quota or credits, even when meta.total is positive. Request-rate limits still apply. For reveal_status=unrevealed, all_revealed: true appears only when the first noncursor page verifies the complete, nonempty match set is already revealed. An empty partial or later page does not establish this. Follow hasMore and the normal pagination fields; an absent hint does not mean unrevealed matches exist.

# Only return entities you've already revealed (an ordinary search)
curl "https://indexed.vc/api/v1/companies?q=fintech&reveal_status=revealed" \
  -H "X-API-Key: idx_YOUR_KEY_HERE"

Filter companies by the tech stack (languages, frameworks, databases) or app stack (SaaS tools they use) with the techStack and appStack params on GET /companies. This is a premium search — richer intent, higher price.

Pricing: premium searches cost 3 credits on the first nonempty page. Pagination within the same filter set is free for 15 minutes via a session token.

A successful paid premium response includes an X-Search-Session header. Echo it back on subsequent requests with the same filters to paginate at zero cost. Any filter change (including adding a country or funding bound) requires a new paid session when results are delivered. An unpaid empty response does not issue a new session token.

# First nonempty page — 3 credits, session token returned
curl -i "https://indexed.vc/api/v1/companies?techStack=React,PostgreSQL&stackMatch=all" \
  -H "X-API-Key: idx_YOUR_KEY_HERE"
# → X-Credits-Cost: 3
# → X-Search-Session: eyJoIjoi…

# Page 2 with session — 0 credits
curl "https://indexed.vc/api/v1/companies?techStack=React,PostgreSQL&stackMatch=all&page=2" \
  -H "X-API-Key: idx_YOUR_KEY_HERE" \
  -H "X-Search-Session: eyJoIjoi…"
# → X-Credits-Cost: 0

Result cap: premium searches return up to 250 matching companies. For very broadly-adopted tools (e.g. AWS, React), the top 250 by relevance are returned. Combine with other filters (industry, country, funding) to narrow the set.

Rate limits are enforced per API key, sliding 60-second window. Limits vary by plan:

Every response includes rate-limit headers. When the limit is exceeded, you'll get a 429 with a Retry-After header.

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 42
X-RateLimit-Reset: 1712345678

Errors return a flat JSON shape with a human-readable error message and a machine-readable code. Validation errors include a details array with field-level messages.

{
  "error": "Credits exhausted. Top up to continue.",
  "code": "CREDIT_LIMIT_REACHED",
  "reason": "credits_exhausted",
  "remaining": 0,
  "top_up_url": "https://indexed.vc/settings?buy_pack=pack_5k",
  "top_up_pack": { "id": "pack_5k", "credits": 5000, "price_cents": 20000 }
}

// Validation errors include per-field details:
{
  "error": "Invalid query parameters",
  "code": "VALIDATION_FAILED",
  "details": [
    { "field": "limit", "message": "Number must be less than or equal to 100" }
  ]
}

The endpoints below cover the core use cases: searching and fetching companies + investors, batch domain resolution, revealing entity depth, and a free usage check.

Enriching a list of domains in Clay? The Clay integration guide has the exact HTTP API column settings and field paths.

List endpoints return paginated results. Use page and limit to navigate. Every list response includes a meta object with total, page, limit, and hasMore.

{
  "data": [ … ],
  "meta": {
    "total": 3124,
    "page": 1,
    "limit": 25,
    "hasMore": true
  }
}

Paid search page size is up to 100; Free is capped at 10 rows and 3 pages. The response's meta.limit reports the effective page size after plan caps. To navigate, increment page until hasMore is false.

Cursor pagination (bulk pulls): for large sequential exports, pass ?cursor= using meta.next_cursor from the previous page. Cursor mode requires a paid plan and uses slug keysets rather than page/sort. Dataset changes can still affect membership; iterate until next_cursor is null.

The API is currently at version v1. We commit to backwards compatibility within a major version: new optional parameters and response fields may be added, but existing fields and behaviors will not break.

Breaking changes ship under a new version path (e.g., /api/v2). We'll announce deprecation at least 6 months before removing a version.

Machine-readable spec: the full OpenAPI 3.1 definition is published at indexed.vc/openapi.yaml — import it into Postman, Insomnia, or your codegen tool.