Y Combinator Companies API — Directory by Batch & Industry

Every company in the YC Startup Directory as one clean JSON record: name, batch, status (Active, Acquired, Public, Inactive), industry and sub-industry, tags, team size, location and regions, website, one-liner, long description, stage, launch date, hiring and Top Company flags, former names and logo. Search by keyword, filter by batch (W24, S23, X25 …), industry, status or region — the same public search index the directory page itself uses, read without a login or key of your own.

Try it in 1 minute:

1. Press Start — the prefilled input searches the directory for "AI agents" and returns the first 100 matches.
2. Seconds later the Output tab shows one record per company: batch, status, industry, team size, location, website, one-liner, tags and the YC page link.
3. Download JSON, CSV or Excel, or add Batches (one per line) to read whole batches — or the whole directory — and narrow by Industries, Status or Regions.

Price: $0.50 per 1,000 companies (one company = one record). A search without matches and a search that could not be run are free.

Use it to:

• Build a lead list of every YC company in a batch or industry — fintech in W24 and S24, say — with website, location and team size, straight into your CRM.
• Keep a startup database or market map fresh: the directory holds about 6,300 companies, and a run per batch reads all of them.
• Find companies that are hiring in a region ("hiring" plus "Europe" or "Remote") for a job board or an outreach campaign.
• Feed an AI agent or research notebook: ask for "climate", "robotics" or "developer tools" and get structured, deduplicated results with descriptions.
• Track a batch over time — status changes (Acquired, Public, Inactive), team-size growth and new launches — on a schedule.

How it works

1. The Actor opens the public directory page (ycombinator.com/companies) once per run and reads the Algolia application id and the search-only key the page embeds. That key is minted for every page load and expires after about a day, so it is never stored — it is read fresh each run. (The index name is decoded from the key's own restrictions; the application id and index name also have built-in fallbacks.)
2. It then queries the same Algolia index the page uses: your Search query as the full-text query, and your batches, industries, statuses, regions, hiring and top-company switches as facet filters. Results come in pages of 100, up to Max results per search.
3. Algolia stops every search at 1,000 hits, which is why Batches are searched one at a time: a batch has a few hundred companies, so a list of batches reads the whole directory. position is the company's place in its search; input says which search it came from.
4. One record per company. Only the company attributes are requested from the index — founder names are never retrieved — and e-mail addresses and phone numbers are removed from descriptions.

Requests go directly to ycombinator.com and Algolia (no proxy, unless a connection drops); a rate limit (HTTP 429) is waited out and retried.

Input

Field	What it does
query	Keywords to search for (names, one-liners, descriptions, tags). Empty = filters only.
batches	Batches, one per line: W24, S23, F25, X25 (Spring 2025) or Winter 2024. Each batch is its own search.
industries	Industry names, one per line: B2B, Fintech, Healthcare, Consumer, Industrials, or a sub-industry such as Infrastructure, Security, Legal. Case does not matter.
statuses	Active, Acquired, Public, Inactive (any of them).
maxResults	Companies per search (per batch), up to 1,000; default 100.
includeDescription	Add the long description (default on).
regions	Region or country names: Europe, United States of America, United Kingdom, India, Remote, Fully Remote …
onlyHiring	Only companies marked as hiring.
onlyTopCompanies	Only companies on YC's Top Companies list.
maxConcurrency	Batches searched in parallel (default 2).
proxyConfiguration	Used only to retry after a network error; the first try is direct.

Output

One record per company:

Field	What it is
id, name, slug, url	Identity and the company's page on ycombinator.com
website, logo	The company's site and small logo
oneLiner, description	The pitch and the long description (contacts removed; description null when switched off)
batch, companyStatus, stage, launchedAt	Winter 2024-style batch, Active / Acquired / Public / Inactive, Early / Growth, launch date (ISO 8601)
industry, subindustry, industries[], tags[]	Classification as the directory files it
teamSize, location, regions[]	People (integer), locations text, region and country names plus Remote flags
isHiring, isTopCompany, isNonprofit, formerNames[]	Flags and earlier names
input, position	Which search the record came from (query and filters in words) and its 1-based position there
status, error, scrapedAt	OK / NO_RESULTS / FAILED, the reason, and when

Example

{
  "input": "\"AI agents\" [batch: Fall 2026]",
  "position": 1,
  "id": 35422,
  "name": "Praxis",
  "slug": "praxis",
  "url": "https://www.ycombinator.com/companies/praxis",
  "website": "https://runpraxis.ai",
  "oneLiner": "AI agents that run law firm operations.",
  "description": "At Praxis, we've built AI agents that run law firm operations, from first inquiry to paid invoice.",
  "batch": "Fall 2026",
  "companyStatus": "Active",
  "industry": "B2B",
  "subindustry": "B2B -> Legal",
  "industries": ["B2B", "Legal"],
  "tags": ["Legal", "LegalTech", "AI"],
  "teamSize": 2,
  "location": "San Francisco, CA, USA",
  "regions": ["United States of America", "America / Canada"],
  "stage": "Early",
  "launchedAt": "2026-10-14T21:59:49.000Z",
  "isHiring": false,
  "isTopCompany": false,
  "isNonprofit": false,
  "formerNames": [],
  "logo": "https://bookface-images.s3.amazonaws.com/small_logos/aac3b9688bfcfe828ca1449c50d9d85d7e55ad89.png",
  "status": "OK",
  "error": null,
  "scrapedAt": "2026-10-19T12:00:00.000Z"
}

FAQ

Is this an official API? No, it's an unofficial API built on public pages — the YC Startup Directory and the public, search-only Algolia index that renders it. The data is exactly what the directory shows; this Actor only shapes it into stable JSON records.

How do I get the whole directory? List every batch in Batches (W05 … S26 and the newer F and X batches) and set Max results to 1,000. Each batch is one search of a few hundred companies, so nothing is cut off. A single unfiltered search stops at 1,000 — that is Algolia's limit, not ours.

Does it include founders? No. Founder names and profiles are not requested from the index and never appear in the output. Descriptions are scrubbed of e-mail addresses and phone numbers.

Why does my batch give NO_RESULTS? Batches are spelled W24, S23, F25, X25 or Winter 2024; a typo such as WI24 matches nothing. Industry and region names must be the directory's own (Fintech, Real Estate and Construction, United Kingdom); case does not matter.

What happens when YC changes the page? The search key is read from the page on every run. If the page stops exposing it, the run ends with a free FAILED record saying so instead of guessing — tell us on the Issues tab and we fix it.

Like it? A short review on the Store page helps other people find this Actor. Something missing or broken? Tell us on the Issues tab — we read every one.

Related Actors

• Hacker News Scraper — stories, comments and search
• Website Tech Stack Detector — CMS, frameworks and analytics of any site
• Google News API — headlines by keyword and topic
• Sitemap Extractor — every URL, date and image from a site's sitemaps
• Article Extractor API — title, author, date and full text of any article