Playbook · Lead Generation

How many employees does a company have? Look it up by name, with the source

Ask Google AI Mode how many employees a company has on LinkedIn and get the size range, the member count when it is given and, usually, the source link, as JSON.

In short

Call the Real-Time Web Search /ai-mode endpoint with a prompt such as "How many employees does Tuple (tuple.app) have on LinkedIn? Reply in one line: employees: <number of employees listed>; size range: <company size range>". reply_parts holds the answer, and reference_links usually holds the source.

The use case

Company size is one of the first filters in account targeting, and private companies rarely publish it. LinkedIn and workforce trackers such as Revelio Labs and PitchBook show estimates. Asking Google AI Mode by company name returns the LinkedIn figures, usually with the link, so you can see where each number came from. For GTM and lead generation teams, it sizes an account list for an ICP filter or a routing rule.

Who does this: SDRs, RevOps teams and GTM engineers sizing accounts for an ICP, recruiters and staffing firms sizing target companies, VC and private equity analysts screening private companies, and market researchers.

How to do it

Uses the Real-Time Web Search API, tested live on November 4, 2026. One call per company. Put the domain next to the name so the answer is about the right company, name LinkedIn as the source, and ask for a one-line format:

import re
import requests

r = requests.get(
    "https://api.openwebninja.com/realtime-web-search/ai-mode",
    headers={"x-api-key": "YOUR_API_KEY"},
    params={"prompt": "How many employees does Tuple (tuple.app) have on LinkedIn? Reply in one line: "
                      "employees: <number of employees listed>; size range: <company size range>"},
)
data = r.json()["data"]
text = " ".join(p.get("text") or "" for p in data.get("reply_parts") or [])
m = re.search(r"employees:\s*([^;]+);\s*size range:\s*([^\n]+)", text)
count = re.fullmatch(r"~?\s*(\d[\d,]*)", m.group(1).strip()) if m else None  # a plain number; "~" is dropped
size = re.search(r"\d[\d,]*\s*[-–]\s*\d[\d,]*|\d[\d,]*\+", m.group(2)) if m else None
print("Employees:", count.group(1) if count else "not found")
print("Size range:", size.group(0) if size else "not found")
for ref in (data.get("reference_links") or [])[:2]:
    print("Source:", ref.get("source"), ref.get("link"))
Result (Python)
Employees: 14
Size range: 2-10
Source: LinkedIn·Tuple https://www.linkedin.com/company/tupleapp
Full response of the cURL call above (Tuple) · JSON
{
  "status": "OK",
  "request_id": "4607008e-549e-44e3-bdfb-4327fbc4af72",
  "parameters": {
    "prompt": "How many employees does Tuple (tuple.app) have on LinkedIn? Reply in one line: employees: <number of employees listed>; size range: <company size range>",
    "gl": "us",
    "hl": "en"
  },
  "data": {
    "reply_parts": [
      {
        "type": "paragraph",
        "text": "employees: 14; size range: 2-10 employees"
      }
    ],
    "reference_links": [
      {
        "title": "Tuple - LinkedIn",
        "link": "https://www.linkedin.com/company/tupleapp",
        "snippet": "View all 14 employees About us Tuple is an app for remote pair programming, optimized for everything that developers care about: s...",
        "source": "LinkedIn·Tuple",
        "favicon": null
      }
    ],
    "session_token": "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"
  }
}

For a whole list

For an account list, one call per company, with the count, the size range, the source and the full answer in a CSV:

import csv
import re
import requests

API_KEY = "YOUR_API_KEY"
URL = "https://api.openwebninja.com/realtime-web-search/ai-mode"
companies = [  # name, domain (the domain tells companies with the same name apart)
    ("Gusto", "gusto.com"),
    ("Clay", "clay.com"),
    ("Mercury", "mercury.com"),
    ("Notion", "notion.so"),
    ("Ramp", "ramp.com"),
    ("Linear", "linear.app"),
    ("Tuple", "tuple.app"),
]

with open("company_headcount.csv", "w", newline="") as f:
    w = csv.writer(f)
    w.writerow(["company", "domain", "employees", "size_range", "source", "source_url", "answer"])
    found = 0
    for name, domain in companies:
        prompt = (f"How many employees does {name} ({domain}) have on LinkedIn? Reply in one line: "
                  "employees: <number of employees listed>; size range: <company size range>")
        r = requests.get(URL, headers={"x-api-key": API_KEY}, params={"prompt": prompt}, timeout=90)
        data = r.json().get("data") or {}
        text = " ".join(p.get("text") or "" for p in data.get("reply_parts") or [])
        m = re.search(r"employees:\s*([^;]+);\s*size range:\s*([^\n]+)", text)
        count = re.fullmatch(r"~?\s*(\d[\d,]*)", m.group(1).strip()) if m else None  # a plain number; "~" is dropped
        size = re.search(r"\d[\d,]*\s*[-–]\s*\d[\d,]*|\d[\d,]*\+", m.group(2)) if m else None
        ref = (data.get("reference_links") or [{}])[0]  # can be empty
        found += bool(count)
        w.writerow([name, domain, count.group(1) if count else "", size.group(0) if size else "",
                    ref.get("source"), ref.get("link"), text.strip()])
print(found, "of", len(companies), "companies returned a plain employee count")
company_headcount.csv
company,domain,employees,size_range,source,source_url,answer
Gusto,gusto.com,,"1,001–5,000",LinkedIn·Gusto,https://www.linkedin.com/company/gustohq,"employees: over 1,000; size range: 1,001–5,000 employees"
Clay,clay.com,"1,580","501-1,000",LinkedIn·Clay,https://www.linkedin.com/company/grow-with-clay,"employees: 1,580; size range: 501-1,000 employees"
Mercury,mercury.com,,"1,001-5,000",LinkedIn·Mercury,https://www.linkedin.com/company/mercuryhq,"employees: Not publicly disclosed as an exact single-digit count on LinkedIn; size range: 1,001-5,000 employees"
Notion,notion.so,800,"501-1,000",LinkedIn·Notion,https://www.linkedin.com/company/notionhq,"employees: ~800; size range: 501-1,000 employees"
Ramp,ramp.com,"2,900","1,001–5,000",LinkedIn·Ramp,https://www.linkedin.com/company/ramp,"employees: 2,900; size range: 1,001–5,000 employees"
Linear,linear.app,,51-200,LinkedIn·Linear,https://www.linkedin.com/company/linearapp,employees: 51-200 (associated members); size range: 51-200 employees on Linear LinkedIn
Tuple,tuple.app,14,2-10,LinkedIn·Tuple,https://www.linkedin.com/company/tupleapp,employees: 14; size range: 2-10 employees

Seven companies took 43 seconds. All 7 returned a size range and cited the company's LinkedIn page; 4 returned a plain member count. Gusto ("over 1,000"), Mercury ("Not publicly disclosed...") and Linear ("51-200 (associated members)") stayed blank, with the full answer kept for review.

Good to know

  • Naming the source in the prompt keeps answers comparable. Without "on LinkedIn", the answer for Mercury listed 1,676 (PitchBook), 1,807 (Tracxn) and 3,200 (Growjo) on November 3, and a 1,001 to 5,000 range citing LinkedIn and Glassdoor on November 4. With it, every answer in our runs cited the company's LinkedIn page.
  • The size range is the steadier column: every company returned one in every run. The member count did not: Gusto came back as 4,797, as "over 1,000" and once as its LinkedIn follower count (167,537) on the same day. The code keeps a count only when the answer is a plain number (a leading ~ is dropped) and leaves the rest blank, with the full answer in the last column.
  • The member count can sit outside the size range. Notion came back with 7,211 members against a 501-1,000 range in two runs (and ~800 in a third), and Tuple with 14 against 2-10. For an ICP filter, the size range is the safer column.
  • reference_links can come back empty: it did in 4 of 47 calls in our runs, so the code prints only the sources it gets.
  • Both LinkedIn figures are estimates: Revelio Labs put Gusto at 4,053 on the same day LinkedIn showed 4,797 members.

Frequently asked questions

How do I find out how many employees a company has?

For a private company, the usual places are its LinkedIn page and workforce trackers. The call above returns the LinkedIn figures with the link, for any company you can name.

Is there an API to get the employee count of private companies by name?

Yes. Send the company name and domain in an AI Mode prompt through the Real-Time Web Search API, as above. Small companies work too: Tuple came back with a 2-10 size range and 14 members.

What are the company size categories based on number of employees?

There is no single standard. The ranges in our answers were LinkedIn's, for example 51-200 for Linear and 1,001-5,000 for Gusto and Ramp. The script stores the range next to the count, so you can filter on either.

How up to date are the employee counts?

Queries run in real time with no caching. The figure is as current as the LinkedIn page or tracker it cites at the time of the call.

Start building with real-time web data

Every API includes a free tier. No credit card required, and the first request is a minute away.