For AI agents & developers

A toolbox for AI agents.

Dealroom maps the world's tech ecosystems — 3M+ startups, investors, founders and funding rounds. This page (and /llms.txt) is the front door for agents: public APIs, machine-readable data, and a chart factory. No key, no sign-up. Point your agent here and let it build.

Paste this to your agent

You have access to Dealroom.co, a live database of 3M+ startups, investors and funding, exposed as agent-friendly tools. Start by fetching https://dealroom.co/llms.txt for the index. Try the public APIs: search market maps with GET https://dealroom.co/api/marketmaps?q=AI, pull live funding with GET https://dealroom.co/api/live-signals, and build a branded chart — GET https://dealroom.co/api/create-chart returns the manual, then POST a spec. Explore, build something, and tell me what worked and what didn't.

Tools you can call

Live JSON/HTML endpoints on https://dealroom.co. No auth, CORS-enabled — call them directly from any client.

GET

/api/marketmaps

Search 170+ curated sector landscapes by keyword or tag. Returns ranked maps with company & segment counts and URLs. Add format=llms for Markdown.

Returns: { count, total, filter, availableTags[], results: [{id, title, description, url, dealroomListUrl, tags, companyCount, segmentCount, companiesUrl}] }. Both camelCase (companyCount) and snake_case (company_count) keys are present.

curl "https://dealroom.co/api/marketmaps?q=fintech&limit=5"

GET

/api/marketmap?id=

Drill into a single map and list its companies (name, website, segment, funding, HQ, founders). Public sample (~12/segment); each segment reports its real total_count.

Returns: { id, title, url, dealroomListUrl, total_companies, returned, capped, segments: [{title, total_count, returned}], companies: [{name, path, dealroomUrl, website, tagline, segment, employees, launchYear, totalFunding:{amount,currency}, hq:{city,country}, industries[], founders[]}] }.

curl "https://dealroom.co/api/marketmap?id=landscape-48993"

GET

/api/live-signals

The latest venture funding rounds — company, city, stage, lead investors and amount.

Returns: { source, fetchedAt, signals: [{type, city, title, meta, value, time}] }. (Note: the array is signals, not results.)

curl "https://dealroom.co/api/live-signals?limit=20"

GET

/api/list-reports

Recent Dealroom research reports with direct PDF links: {id, title, slug, date, pdfUrl}.

curl "https://dealroom.co/api/list-reports"

GET

/api/third-party-maps

Community- and partner-built market maps, with source attribution and company counts.

Returns: { count, results: [{slug, title, description, source_label, source_url, public_url, tags, company_count, category_count}] }. Attribution is in source_label (e.g. "Extantia"); source_url may be null.

curl "https://dealroom.co/api/third-party-maps"

GET

/api/dashboard-charts

The catalogue of charts that already exist on Dealroom dashboard pages (/multiples/, /power-law-investor-ranking/, /vc-investment-dashboard/). Each entry: id, title, kicker, description (what story it tells), snapshotUrl, and the filterParams the chart accepts via URL. This inventory powers two equally valid agent workflows: (A) Story-first / content-marketing — you have an angle from the news or a fresh data signal, and you scan this list to see whether an existing chart already tells it; (B) Asset-first / product-marketing — you start by browsing this inventory, pick an interesting chart, and frame a story around what the data shows today. Alternate between the two. Either way, if a dashboard chart matches your angle, capture it via window.__chartSnapshot() rather than rebuilding — page-level charts have bubble groupings, company logos in tiles, country flags and other hand-tuned details the API can't reproduce.

Returns: { description, captureContract, referencePattern, count, countByPage, charts: [{id, page, title, kicker, description, filterParams, filterValues, snapshotUrl}] }.

curl "https://dealroom.co/api/dashboard-charts"

GET

/api/lookup-slug

Resolve a free-text company / investor name to its canonical Dealroom slug, with the ready-to-use URLs for entity-news, sentiment and the Markdown twin attached to each match. Call this before /api/entity-news or /api/sentiment so the slug actually matches.

Returns: { query, type, count, totalMatches, matches: [{name, slug, type, score, newsUrl, sentimentUrl, markdownUrl, profileUrl}], sources[], hint }. Matches are sorted by score (exact > compact > prefix > substring); hint populates only when nothing matches.

curl "https://dealroom.co/api/lookup-slug?name=Anthropic&type=company"

GET · POST

/api/create-chart

GET returns a self-describing manual — every chart type, every field, and ready-to-POST example payloads. POST a spec and get a standalone, on-brand Dealroom chart as HTML; add ?format=json to get {id, url, html} back (the url is the hosted chart, also in the X-Chart-Url header). Start with the GET.

# 1. read the manual (chart types, fields, examples)
curl "https://dealroom.co/api/create-chart"

# 2. make a chart, get back {id, url, html}
curl -X POST "https://dealroom.co/api/create-chart?format=json" \
  -H "Content-Type: application/json" \
  -d '{"chartType":"bar","headline":"Top 5 countries by VC funding",
       "labels":["US","UK","France","Germany","Sweden"],
       "datasets":[{"label":"Funding","data":[120,21,13,9,7]}],
       "unit":"$B","source":"Dealroom.co"}'

POST

/api/agent-feed/<id>

Update, publish, or delete a chart you made. The action goes in the JSON body, not the query string. delete removes it from D1 + KV (the public URL 404s within a few minutes — note the 5-min edge cache); publish / unpublish toggles visibility on the Resources gallery; regenerate re-renders against the current template code.

curl -X POST "https://dealroom.co/api/agent-feed/<id>" \
  -H "Content-Type: application/json" \
  -d '{"action":"delete"}'

Also available (beta — use the exact Dealroom slug): /api/sentiment?entity=Spotify (AI-synthesized X sentiment) and /api/entity-news?path=spotify&type=company (latest news; type=investor for funds). Don't guess slugs — call /api/lookup-slug?name=<query> first; it returns the canonical slug plus the ready-to-call URLs for both. As fallbacks, the slug is also the trailing path on app.dealroom.co/companies/<slug>, the companies[].path field from /api/marketmap, or any entry in /investors/manifest.json.

Make a chart, get an image

Every chart from /api/create-chart renders the Dealroom design system (typography, palette, logo) from a single contract — so an agent gets publication-ready output, not a generic bar chart.

Pick a chart type — 15 standard + 13 specialty types (market map, treemap, scatter, choropleth, marimekko, bubble…). The GET manifest lists them with examples.
Pass bare: true in the spec for a chrome-free export — hides the theme toggle and PNG button so the rendered HTML is pure card, ideal for headless capture.
Capture the PNG from a headless browser — load the returned url in Puppeteer/Playwright, wait for window.__chartReady, then call window.__chartPng({scale: 3}) in the page. For iframe pipelines, append ?autoPng=1 to the URL and listen for postMessage({type:'chartPng', dataUrl}). Full guide: /docs/agent-chart-png-guide.md.
Embed in your own page — every chart honors the 5-slot frame embed contract (fluid responsive CSS, intrinsic sizing, postMessage theme handoff).
Design in the browser — the Chart Studio is the visual editor on top of the same API.

// Puppeteer / Playwright: produce a chart, then capture its PNG
const r = await fetch('https://dealroom.co/api/create-chart?format=json', {
  method: 'POST', headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({ ...spec, bare: true })   // bare:true = chrome-free export
}).then(r => r.json());

await page.setViewport({ width: 1600, height: 900, deviceScaleFactor: 1 });  // pick aspect per channel
await page.goto(r.url);                            // r.url = /api/serve-chart/<id>
await page.waitForFunction('window.__chartReady === true');
const { dataUrl } = await page.evaluate(() => window.__chartPng({ scale: 3 }));
// dataUrl is the base64 PNG, ready to write or post.

Common recipes

End-to-end flows that combine the endpoints above — start from a story angle, the data follows.

Sector landscape post

Find a sector landscape, list its companies, chart them.

# 1. discover the landscape
curl "https://dealroom.co/api/marketmaps?q=fintech&limit=5"

# 2. drill into one — pick an id from results[]
curl "https://dealroom.co/api/marketmap?id=landscape-47614"

# 3. compose a marketMap or treemap from companies[] grouped by segment, then POST it
curl -X POST "https://dealroom.co/api/create-chart?format=json" \
  -H "Content-Type: application/json" -d '{...}'

Funding hook

Pull the day's venture rounds, chart the largest by city, sector or stage.

# 1. fetch live rounds
curl "https://dealroom.co/api/live-signals?limit=20"

# 2. bucket signals[] by city or stage, then POST a horizontalBar or treemap
curl -X POST "https://dealroom.co/api/create-chart?format=json" \
  -H "Content-Type: application/json" -d '{...}'

Resolve a slug, grab news + sentiment

Turn a free-text name into the canonical slug, then chain the news and sentiment endpoints — the lookup response hands you the URLs.

# 1. resolve "Anthropic" → matches[].slug, .newsUrl, .sentimentUrl
curl "https://dealroom.co/api/lookup-slug?name=Anthropic&type=company"

# 2. follow the URLs the lookup returned
curl "https://dealroom.co/api/entity-news?path=anthropic&type=company"
curl "https://dealroom.co/api/sentiment?entity=Anthropic"

Investor profile post

Resolve an investor slug, read the Markdown twin, optionally chart their stage breakdown.

# 1. list all investors with rank + slug
curl "https://dealroom.co/investors/manifest.json"

# 2. read the Markdown twin — LLM-friendly metrics + portfolio shape
curl "https://dealroom.co/investors/sequoia-capital.md"

# 3. (optional) chart the investor's stage / sector breakdown via /api/create-chart

Machine-readable data & discovery

Where an agent looks first, and the structured data behind every page.

/llms.txt — the master index: tools, market maps, investors, rankings and multiples.
/data/just-founded.json — the weekly Just Founded feed: newly founded startups with founders, photos, a website screenshot, and ready-to-post social.post + social.hook copy. Filter on fundingStatus.
/sitemap.xml — sitemap index (companies, investors, countries, cities, regions, sectors, universities, landings).
/charts/landscapes-manifest.json — every market map with tags, counts and URLs.
/investors/manifest.json — investor profiles; each has a Markdown twin at /investors/{slug}.md (e.g. sequoia-capital.md).
/resources/multiples-data.json — the EV/Revenue multiples dataset (4,581 exit & VC rounds).
Flagship pages (multiples, Power Law ranking) ship Schema.org Dataset, ItemList and FAQPage JSON-LD.

Welcoming to AI crawlers by design — see /robots.txt. This is beta and evolving; if something's missing or broken, that's useful feedback.