Data & API
Everything on Venture Atlas - 97 companies across 16 industries, 426 dated upcoming milestones, the news digest, and the sector briefs - is published in machine-readable form. Free, no API keys, CORS-open, and derived from the same data as the pages at every deploy, so it is never staler than the site. Every field carries the source URL it came from. Last data update: 2026-07-29.
/ JSON API
Four read-only endpoints, prerendered as static files on the CDN. Responses share an envelope with the API version and the date of the last data update. There are no query parameters; pick the endpoint with the scope you need.
Every tracked company as a slim record: identity, industries, parent, milestone track record, data-verified date, and pointers to the full profile.
One company in full: key metrics, dated next steps, resolved milestone history, timeline, funding rounds and total, products, subsidiaries, and its news.
The industry map with member companies and the full State of the Industry briefs.
Every dated, still-upcoming milestone across all companies, soonest first, each with its source and the company's track record.
/ MCP Server
An assistant with Model Context Protocol support can query the dataset live as tools instead of scraping pages. The server is public, unauthenticated, and stateless, served over Streamable HTTP at /api/mcp. To add it to Claude Code:
claude mcp add --transport http venture-atlas https://www.ventureatlas.org/api/mcpFind companies by free text and/or industry; returns slim records with slugs.
The full profile for one company, including the long-form overview on request.
Announced, still-upcoming milestones with target dates, soonest first.
The daily news digest, filterable by company or date.
Any other MCP client works with the same URL over the HTTP (not stdio) transport. Listed in the official MCP registry as org.ventureatlas/venture-atlas. A plain GET on the endpoint returns a description of the server, its tools, and current dataset counts. Rate limit: 120 requests per minute per IP.
/ Plain Text & Feeds
For assistants and crawlers that read text rather than JSON.
A curated plain-text map of the site for assistants and answer engines, per llmstxt.org: every industry and company with a one-line summary.
The full facts layer as plain text: sector briefs, every company's metrics, dated next steps, track record, and funding, each line with its source URL.
RSS feed of the daily news digest, newest 50 items.
Every page on the site, with per-page data-freshness dates.
/ Using the Data
The dataset is curated by Venture Atlas and free to use. Every fact we publish links to the press release, filing, or outlet it came from - when you republish a claim, cite that primary source; when you use our curation, a link back to the company page it came from is appreciated. All surfaces on this page are rebuilt from the same committed data on every deploy, so they never disagree with the site or each other.