# Company data API for AI agents: from $0.001 per call ($1 per 1,000), no subscription

> Whizbrew's Company data API costs from $0.001 per call ($1 per 1,000 calls); a typical call is $0.004 ($4 per 1,000). Only a call that returns data is charged: an empty or failed call is free. There is no subscription and no monthly fee. A new account starts with $2 of free credit, with no card.

It reads a company's profile, facts, team, pricing page, tech, news, hiring and competitors, from its own site and public registers: 34 tools, live at the moment of the call, over one API key or one MCP server (`https://whizbrew.com/mcp`). It is an API, not a scraper you run: no proxies, no browser and no scraping code on your side. The same key reaches 893 tools for other platforms ([all APIs](https://whizbrew.com/api), [all tools](https://whizbrew.com/tools?a=2ffbdd2c26e3), [pricing](https://whizbrew.com/pricing?a=2ffbdd2c26e3)).

## Pricing

| Tool | What it returns | Per call that returns data | Per 1,000 calls |
|---|---|---|---|
| `company.legal_entity` | A company's legal identity from the global LEI register: legal name, addresses, registration number and status. By LEI it also gives the ultimate parent. | $0.001 | $1 |
| `company.registry_search` | Search an official public company register by company name: France (fr), Norway (no), Finland (fi), Czech Republic (cz), Estonia (ee) or the United Kingdom (uk, read from the register's public search page). One row shape for all: name, registry number, active or not, legal form, founding date, address, industry code, staff size where the register gives it, and the register's page for the company. Finland and Estonia answer one page only; a Czech search with over 1,000 matches is refused. Nothing is charged when nothing is found. | $0.002 | $2 |
| `company.registry_record` | One company from an official public company register by its registry number: France (SIREN), Norway (organisation number), Finland (business ID), Czech Republic (IČO), Estonia (registry code), United Kingdom (company number) or Poland (KRS number). The same row as the search, plus what the register publishes openly: officers for France, Norway and the Czech Republic, share capital for Norway, the Czech Republic and Poland, last published revenue and net income for France. Poland hides the names of people; Finland, Estonia and the UK search page publish no officers here. | $0.002 | $2 |
| `company.newsroom` | Where a company publishes its news: the newsroom or press page, blog, investor relations, careers, changelog and status page and the RSS/Atom feeds, found from the links of the company's home page; for kinds it does not link, up to seven usual paths (/press, /newsroom, /blog, /careers, /investors, /feed, /rss.xml) are tried. Each result says whether it was linked, declared or guessed. Reads the home page only; a site that refuses automated readers is an error. Nothing found is free. | $0.003 | $3 |
| `company.from_domain` | Who is behind a domain: the organisation the home page itself names (its structured data: name, legal name, logo, profiles, address; the site name; the copyright line) and the Wikidata entities whose official website is this domain (name, description, country, founding year, stock ticker). Every candidate says where it came from; best_name is the strongest one. Nothing is inferred from the domain name: no evidence gives an empty list, which is free. | $0.002 | $2 |
| `company.domain_from_name` | The website of a company, from its name: candidates from two public sources, the entities with an official website in Wikidata and the web search results whose host name is the company name, each with its sources, a score and a confidence (high, medium, low), plus description and search position where known. best_domain is set from medium confidence on. Nothing is made up from the name: no name + .com is tried. A name that matches nothing is free. | $0.002 | $2 |
| `company.key_pages` | Which of the usual pages a company's website has, from the links of its home page: pricing, about, team, customers, integrations, changelog, status page, careers, blog, docs, contact, press, events, partners, locations, security, privacy, terms, log in, sign up and demo, each with its address and the link's text; the kinds the home page does not link are listed too. One request to the site. Use it to see what a company publishes before asking for one page (company.pricing, company.team ...). | $0.002 | $2 |
| `company.pricing` | The plans and prices on a company's pricing page: each plan with its name, price (amount, currency, as written), what it is charged per (user, month ...), billing note, description, feature list and button; plus whether there is a free plan, a free trial (and its days), monthly or annual billing, the annual saving, contact-sales and usage-based pricing, the currencies, and every price on the page with its line (also where no plan was recognised). The page is found from the home page's links, else at /pricing or /plans, else the plans on the home page itself. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. Prices shown only after a click (a yearly switch, a currency menu) are not seen. A site with no pricing page is free. | $0.004 | $4 |
| `company.team` | The people a company names on its own team, leadership or about page: name, role, the group they stand under (Leadership, Board, Investors ...) and the profile links the page shows next to them (LinkedIn, X, GitHub). Only what the company itself publishes: nobody is looked up anywhere else, and no contact detail is added. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A person is a name next to a job title, so a page of photos without titles gives little. A site that names nobody is free. | $0.004 | $4 |
| `company.public_contacts` | The contact details and people a company itself prints: email addresses from its website, its press page and its security.txt (each marked as a role mailbox like press@ or a named mailbox, and whether it is on the company's domain), phone numbers, postal addresses, whether there is a contact form, and the people its team page names with their roles. Each entry says where it stands. Public evidence only: no address is guessed or completed (email.guess does that and says so) and nobody is looked up in any directory or database. A site that prints nothing is free. | $0.006 | $6 |
| `company.people` | People at a company, optionally by job title ("product manager", "vp sales"): everyone the company itself names in public, merged from its team or leadership page, the speakers of its events and the authors of its blog, each with the role as published, profile links the page shows, which sources name them and the page they were read from. Public evidence only: nobody is looked up in a people database and no email or phone is added, so a small or mid-size company is covered well and a large one shows its leadership and public voices, not its whole staff. "title" keeps the people whose published role contains every word. A company that names nobody (or nobody with that title) is free. | $0.006 | $6 |
| `company.blog_authors` | Who writes a company's blog, from its own feed: each author the feed names with the number of posts, share, newest post (title, link, date) and oldest post in the feed. The feed is found from the blog page or the home page (or give feed). A feed holds only recent posts, so this is a sample of recent writing, and many feeds name no author: such an answer is free. | $0.003 | $3 |
| `company.customers` | The customers a company shows on its website: the names on the logos of its customers page (and the logo wall of its home page) and its case studies, each customer with what names it (logo, case study) and the links; plus the numbers the company states about its customers ("Trusted by 50,000 companies"). These are the customers the company chooses to show, not its customer list. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A logo is named by its alt text, so pictures without one give no name. A site that shows none is free. | $0.004 | $4 |
| `company.integrations` | The products a company says it integrates with, from its integrations, apps or marketplace page: each integration with its name, its page, category and short description where the page gives them, and the categories of the directory. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A directory that loads more as you scroll shows only its first screen here. A site with no such page is free. | $0.004 | $4 |
| `company.changelog` | What a company says it shipped, from its changelog, release notes or "what's new" page: the entries with date, title, link and first lines, the newest and oldest date on the page, days since the newest, how many entries fall in the last 30 and 90 days and the typical gap between entries: a plain measure of how actively the product ships. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. Only what the one page shows is counted. A site with no such page is free. | $0.004 | $4 |
| `company.status` | A company's public status page: where it is, the state now (operational, degraded, partial_outage, major_outage, maintenance), the open incidents and planned maintenance, and for the common status page systems every component with its state, the newest incidents (name, impact, start, resolution, minutes to resolve) and how many incidents the last 30 and 90 days had. Found from the home page's links or at status.<domain>; other kinds of status page give the state from their words. A company with no public status page is free. | $0.002 | $2 |
| `company.careers_page` | A company's careers page as it stands on its own site: where it is, the hiring system behind it (Greenhouse, Lever, Ashby, Workday ... with the tool that reads the full board), the job links the page itself shows (title, place, link, the team heading above), whether it says there are no open roles, the benefits it lists and whether it mentions remote or hybrid work, equity, visa sponsorship and internships. For every open job use jobs.company_board; this tool is the page and its message. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A site with neither a careers page nor a job board is free. | $0.005 | $5 |
| `company.locations` | The offices and addresses a company publishes: from its locations or offices page (a place name with the address under it), the organisation data in its home page's markup, and its contact, about and imprint pages. Each address says which of the three it came from and the page; the same address found twice is listed once. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A company that names only cities or shows a map gives nothing here, and that answer is free. | $0.005 | $5 |
| `company.legal_pages` | A company's legal and trust pages, from the links of its home page: privacy policy, terms, cookies, data processing agreement, subprocessors, imprint, acceptable use, SLA, GDPR/CCPA, accessibility, refunds, vulnerability disclosure and security or trust centre. Each with its address, title, length and the date the page itself gives as last updated or effective (read from up to 8 pages). Privacy and terms are also looked for at their usual addresses, and reported when missing. Good for vendor checks and for seeing when a company last changed its terms. | $0.006 | $6 |
| `company.enrich` | One company, one profile, from its website (or its name); the tool to ask for a company's headcount (employees on LinkedIn, size range, open jobs): name and description, logo and phone as the site states them, industry, founding year, size signals (employees and followers on LinkedIn, open jobs), headquarters and addresses, legal name and LEI record, stock ticker, social profiles, technology by category, hiring (system, open jobs by team and place, new in 30 days), recent news and funding headlines, the ad libraries that know it, traffic rank, backlink summary, apps in the stores, key pages (pricing, careers, blog, docs, status ...) and the domain's registration; and, under facts, what registers and the company itself state, each entry with its source and date: reported revenue, net income and R&D with the fiscal period (US regulator filings, listed companies), industry code, officers and directors, private fundraising notices with date, amount and number of investors and their sum (US Form D), employees and revenue with their year, headquarters, parent, subsidiaries, chief executive and founders (Wikidata's item for the website), the LEI record, the company register's record (UK, France), the figures the company states on its about page and the mailboxes and team its own site prints. Composed live from about 20 of our own tools at the same time, inside one time budget; sections names the tool behind each part and whether it answered, partial says when one did not. Facts are tied to the company by its domain; what is matched by name (news, funding, app search, ad libraries) says so. Nothing is estimated and no contact database is used: a field that is absent was not found. One price for the whole profile; a domain with nothing behind it is free. | $0.025 | $25 |
| `company.enrich_bulk` | company.enrich for up to 5 websites in one call, all at the same time: one profile per domain in the order given, each the same shape as company.enrich. A domain that cannot be read has error and message in its place and does not fail the call. One price for the call, whatever the number of domains (5 at most); charged when at least one profile has data. | $0.10 | $100 |
| `company.news` | The recent news about a company in one list: our two news searches, PR Newswire releases, recent funding headlines and the company's own news or blog feed, merged newest first with one row per story (found_by names every source that had it), each marked news, press_release, funding or own_post. An outside item is kept only when the company's name stands in its headline or snippet; for a name that is also a common word add context. A company with no recent news is free. | $0.006 | $6 |
| `company.social_stats` | A company's followers on every platform, in one call: the profiles its own website links to (X, Instagram, TikTok, YouTube, LinkedIn, Facebook, GitHub, Threads, Bluesky, Pinterest, Twitch, Telegram and more) are each read now from the platform's public page, and returned with followers, the largest platform and the total. Only profiles the site itself links or names in its organisation data are read: no handle is guessed. A site that links no profile is free. | $0.01 | $10 |
| `company.app_presence` | A company's apps on the stores we read: the store pages its own website links to (exact), and the apps a search for its name finds in the App Store, Google Play and the Microsoft Store whose developer is that name. Each app with store, id, name, developer, rating, number of ratings, installs where the store shows them, price and link, and how it was matched. A company with no app is free. | $0.006 | $6 |
| `company.reviews` | Every review source we can read for a company, in one answer: the open business review sources (Google Maps, the Bing and Apple map listings, the review site keyed by website, the rating the company's own site states) and the ratings of its apps in the App Store and Google Play. Each row says what the rating is about, the score and its scale, the number of reviews, how it was matched and which tool reads the reviews themselves. Ratings are not merged into one score. A company with no rating anywhere is free. | $0.008 | $8 |
| `company.similar` | The companies that public pages put next to this one, each with its evidence: the competitors the company names on its own comparison pages ("X vs Y", "Y alternative"), the other side of web search results titled "X vs Y", and the pages LinkedIn shows next to the company's page (the first six of those also with the website domain their LinkedIn page states). Ordered by evidence, the company's own comparison pages first. No market database is used, so a company nobody compares in public has few or none: that answer is free. | $0.006 | $6 |
| `company.competitors_ads` | Where a company's competitors advertise: for up to 4 competitors (given by name or website, or found from the company's own comparison pages when none is given) the ad libraries that know each one (Google, Meta, Apple App Store, Microsoft, Snapchat, Amazon, Pinterest, TikTok), the matching advertisers, the number of ads where the library gives one and the tool that lists the ads. A library asked by name can match another company of the same name: the matches are listed so you can check. Free when no competitor is found in any library. | $0.02 | $20 |
| `company.pricing_changes` | How a company's pricing changed: its pricing page as the Internet Archive saved it near a past date, compared with the page today (or a second date). Plans added and removed, price changes per plan with the percentage, features added to and removed from a plan, and a free plan or trial that came or went; both copies are named with their capture time. Plans are matched by name. Free when no earlier copy exists or neither copy shows plans. | $0.006 | $6 |
| `company.hiring_trend` | Whether a company is hiring more or less than before, a sign of how its headcount moves: the open roles its job board or careers page listed in the copies the Internet Archive saved about 2 years, 1 year, 6 and 2 months ago, and today, with the change and a plain trend (growing, steady, shrinking). The board is the one the site links, else the board named like the company on the hiring systems we read (kept only when it names the company). Counted from the total the board states itself, else from the job links of the page, so it works for boards and careers pages that list their jobs as links and not for ones that load them with a script; such a company gets no trend and the call is free. For today's exact numbers by team use jobs.hiring_summary. | $0.008 | $8 |
| `company.tech_changes` | Which tools a company's website added and dropped: the technologies our web.tech rules find in the markup of its home page, in the copy the Internet Archive saved near a past date and today (or a second date): analytics, tag managers, ad pixels, chat, marketing and email tools, CMS, frameworks. added, dropped and kept, with both copies named. Only marks in the page's HTML are compared (headers and cookies are not in an archived copy), and the two copies are compared, not every date between. Free when there is no earlier copy. | $0.005 | $5 |
| `company.about` | What a company states about itself on its about page and home page: the year it says it was founded, the size of its team, money raised, investors, customer numbers and reach (countries, offices), each quoted with the line it stands in and the page; plus the page's opening paragraph and headings. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. These are the company's claims, not checked figures. A site that states none is free. | $0.004 | $4 |
| `company.compliance` | The security and privacy frameworks a company says it meets, from its security or trust page and home page: SOC 2 (with the type when stated), ISO 27001 and its relatives, GDPR, CCPA, HIPAA, PCI DSS, FedRAMP, CSA STAR, TISAX, the Data Privacy Framework and more, each quoted with the line it stands in; the practices the page describes (penetration tests, SSO and SAML, encryption at rest and in transit, data residency, audit logs), its trust centre, bug bounty, security.txt and the related legal pages. For vendor checks. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. A named framework is the company's claim, not a certificate we have seen. A site that names none and has no security page is free. | $0.004 | $4 |
| `company.events` | A company's events and webinars, from its own events page: each dated event with title, link and first lines, how many are still to come, and the speakers the page names with their job titles. Read from the company's own website (its pages as our web.read reads them; robots.txt honoured), reported as the company publishes it and found with rules, not a model. Speakers are the people the company itself lists there; nobody is looked up anywhere else. A site with no events page, or one that loads its events with a script, is free. | $0.004 | $4 |
| `company.vat_check` | Is an EU VAT number valid: asks VIES, the European Commission's VAT Information Exchange System, which asks the member state's tax register, and returns valid or not, with the registered name and address where that member state discloses them (Germany and Spain do not: null). Give the number with its country code (IE6388047V) or the number and the country; spaces and dots are ignored. EU member states and Northern Ireland (XI) only: no UK, Swiss or Norwegian numbers. "Not valid" is an answer and is charged; a member state's register being busy or down is an error and is not. One read of the public service, no login, no key, no database of ours. | $0.001 | $1 |

## Example calls

`company.legal_entity` ($0.001 per call): A company's legal identity from the global LEI register: legal name, addresses, registration number and status

```
curl -s -X POST https://whizbrew.com/v1/call/company.legal_entity \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"limit":3,"query":"Spotify"}'
```

`company.registry_search` ($0.002 per call): Search an official public company register by company name: France (fr), Norway (no), Finland (fi), Czech Repu

```
curl -s -X POST https://whizbrew.com/v1/call/company.registry_search \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"country":"fr","query":"doctolib","limit":3}'
```

`company.registry_record` ($0.002 per call): One company from an official public company register by its registry number: France (SIREN), Norway (organisat

```
curl -s -X POST https://whizbrew.com/v1/call/company.registry_record \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"country":"no","number":"923609016"}'
```

Over MCP the same call is `call_tool` with `{"tool": "<name>", "arguments": {…}}`. How often each tool works, measured from real calls: [https://whizbrew.com/status](https://whizbrew.com/status).

## Start

- **An agent on its own**: read [https://whizbrew.com/agents.md?a=2ffbdd2c26e3](https://whizbrew.com/agents.md?src=body-text&a=2ffbdd2c26e3). It asks for a free quote (`GET https://whizbrew.com/v1/agent/guide?goal=<the task in plain words>`), then one request (`POST https://whizbrew.com/v1/agent/start?a=2ffbdd2c26e3`) gives it a key that works at once with the free credit, with no email and no sign-up form. Its human claims the account later.
- **A chat app** (ChatGPT, Claude, Gemini): add the MCP server `https://whizbrew.com/mcp` as a custom connector and sign in. No key to copy.
- **An agent with a shell** (Claude Code, Codex, Cursor): read [https://whizbrew.com/llms.txt?a=2ffbdd2c26e3](https://whizbrew.com/llms.txt?src=body-text&a=2ffbdd2c26e3). The agent starts a pairing with one request, its user approves with one click, and the agent holds its own key.
- **Plain HTTP**: create an account at [https://whizbrew.com/app#signup](https://whizbrew.com/app#signup), make a key, then `POST https://whizbrew.com/v1/call/<tool>` with `Authorization: Bearer <key>`.
