Crunchbase Pro vs API: Pay by the Seat or by the Call

Look up a company's funding rounds without buying a seat. Paste this into your agent:

Set up https://monid.ai/SKILL.md, and then use Monid to search Crunchbase for Stripe, list its funding rounds, and tell me the round count and the type of each round.

One prompt resolves a company and reads its rounds. 1 min setup · 6-14 hrs saved.

You are deciding how to pay for Crunchbase data: a Pro seat for a person, or a call each time a pipeline needs a company. The two are priced on different logic, a fixed fee for unlimited browsing against a small fee per lookup, so the right one depends on who asks the questions and how many. This guide sets the two against each other, shows what a call returned for Stripe on 2026-10-16, and works out where the break-even sits. It runs through Monid, the OpenRouter for agent tools: one key, one balance, every tool priced per call.

What is the easiest way to get Crunchbase data without a Pro seat?

Monid. Install it once, add a key, and your agent calls Crunchbase endpoints directly: resolve a name to an organisation, then read the funding rounds for it. There is no seat to assign and no contract to sign first.

• Pay per call. A search is $0.02 and a funding rounds call is $0.02 (Verified 2026-10-16 on the company enrichment tools page). Buy calls as you need them: no monthly plan and no minimum top-up. Discovery and inspection are free; only a run bills.
• One install, every tool. The same key and balance call Crunchbase, other funding sources and every other tool on Monid.
• Lives inside your agent. Install the CLI, add your key, run the prompt above, and the endpoints are there whenever your agent needs them, with nothing to keep running. Your agent reads SKILL.md and learns the rest, so setup is about a minute.

Looking up one company the way this post does, a search then a rounds call, costs $0.04 (Verified 2026-10-16).

Should you pay by the seat or by the call?

Pay by the seat when people browse the database, and by the call when a pipeline looks up companies you already have. A seat is a fixed fee that makes every search free at the margin. A call is a variable fee that makes every lookup cost the same, whether it is your first or your ten-thousandth.

What does a seat buy?

Database vendors often quote by the seat, on an annual term, for one named person. Inside the seat sits the interface: search across the whole database, saved searches and alerts, list building and exports. You pay for the right to ask as many questions as you like, and the cost does not move with the number of questions.

What does a call buy?

One answer to one question. A search resolves a name, a rounds call lists the rounds for the organisation that search returned. Nothing is assigned to a person, nothing renews, and the bill follows what you ran.

Where is the break-even?

Divide the seat price by the cost of a lookup. A company lookup here is two calls, so $0.04 (Verified 2026-10-16). Every $100.00 of balance buys 2,500 of them (Verified 2026-10-16). If a seat costs you more per year than the number of companies you would look up times $0.04, calls are the cheaper way to ask those questions. If you would browse rather than look up, count the hours people spend in the interface instead, because a loop of calls cannot do that job at any price.

That arithmetic is the whole decision, and it has one condition people skip: the two products do not return the same thing. The next sections show what a call returns, so you can tell whether the lookups you have in mind are the kind a call can answer. The same seat-versus-call split for a different database vendor is in the PitchBook API guide.

📖 See also Startup funding data by API: Crunchbase alternatives that bill per call

How do you look up funding rounds for a list of companies by the call?

Resolve each name to an id, list the rounds for that id, and loop the pair over your list. Forty accounts is eighty calls, $1.60 in total (Verified 2026-10-16), billed as you run them. The table says which route fits which job.

For agents

Grab an API key at app.monid.ai, then paste this to your agent and hand it the key:

set up https://monid.ai/SKILL.md

It learns the whole discover, inspect, run workflow itself. More in the agent quickstart.

For humans

npm install -g @monid-ai/cli
monid keys add -k <your-api-key> -l main

Step 1. Resolve the name

monid run -p crunchbase -e /search_organizations --query '{"query": "Stripe"}'

The response carries count and a page of entities, each with an identifier that holds a uuid and a permalink. Keep both: the profile call wants the permalink and the rounds call wants the uuid.

Step 2. Read the profile if you need it

monid run -p crunchbase -e /get_organization_profile --query '{"slug": "stripe"}'

This takes the permalink from step 1. If the search row already carries what you need, skip it: it is a second billed call.

Step 3. List the rounds

monid run -p crunchbase -e /get_company_funding_rounds --query '{"org_uuid": "<uuid>", "limit": 25}'

The response carries a count and a page of rounds. To run a list, hand the loop to your agent:

📖 See also B2B data provider: dataset or API

Can you export from Crunchbase Pro or scrape it with Python?

You can do both, and each fits one kind of work. A seat usually includes export to a spreadsheet, and for a one-time list it is the quickest route: build the search in the interface, export, done. Export limits differ by plan, and an export is a snapshot a person takes, not a feed a system reads.

Scraping fails in a predictable order. A database sold by the seat sits behind a licensed login, so a scraper needs an account, a session that survives and pages rendered for a signed-in user. Automated extraction is the kind of thing database terms tend to restrict, and the rules differ by vendor and use, so read yours. Then comes the upkeep: layouts change, sessions expire, and matching a scraped name to the right company is a repair job of its own.

For anything you will run more than once, Monid is the better pick. The collection is maintained for you, the output is typed JSON instead of HTML you parse, and you pay for the calls you run, not for a seat nobody opens. The three commands above are the whole integration.

What does one call return compared with what a seat shows?

A call returns the shape of a company and the shape of its funding history. We ran the endpoints on Stripe on 2026-10-16, so everything below is what came back.

The company row

/search_organizations with query Stripe and limit 5 returned count: 70 and five entities in about 16 seconds. Each entity carried seven properties: identifier, short_description, rank_org, categories, location_identifiers, num_employees_enum and status. For Stripe itself that was a rank_org of 86, a size band of c_05001_10000 and a status of operating.

The round list

/get_company_funding_rounds with the uuid from that row and limit 5 returned count: 24 and five rounds in about 22 seconds. Each round carried three properties:

{
  "identifier": { "value": "Series H - Stripe" },
  "investment_type": "series_h",
  "funded_organization_identifier": { "value": "Stripe", "role": "investee" }
}

The types on the page were three secondary_market, one series_h and one series_b. A preview_properties block marked announced_on and money_raised as present on some rounds, with no values in it.

What that means for the seat

Everything a lookup pipeline needs to classify a company came back: its category, size band, status, how many rounds it has raised and what kind they were. The amounts, dates, valuations and investors did not. Those are the fields a person reads on a seat, and they are the reason a seat exists. Write your schema against the three properties, not against the field list a vendor shows in its interface.

You can build three things on the shape alone:

• A funding trigger for outbound. A company with a late-stage round type is growing and has budget.
• A stage filter. Keep accounts with a series_ round and drop the ones with none, before anyone spends time on them.
• A round counter. Store count per company each month and watch it change.

What goes wrong when you replace a seat with calls?

Four things, all from the 2026-10-16 runs. None of them is a reason to avoid calls; each is a thing a seat hides from you.

Why does one name match 70 organisations?

Because name search is fuzzy and a short name is everywhere. Stripe returned count: 70, and the first five were the payments company, a fashion retailer in Devon, a reconciliation startup marked closed, a chai brand marked was_acquired and a financing arm. On a seat, a person glances at the logo. In a pipeline you pick by rank_org or by permalink, never by position alone. Finding a company's website from its name shows the same resolve-first habit.

Why are the calls slow?

Because these are slow endpoints. The search took about 16 seconds and the rounds call about 22, and monid inspect lists a 95th percentile of about 33 seconds for search. A seat answers in the interface almost at once. Run a call-based job as a batch with a few requests in flight, and do not put it in front of a user who is waiting.

Why does count not match the array?

Because limit sizes the page, not the total. The rounds call said count: 24 and returned the five we asked for. Read count for the total, and raise limit only to the number of rounds you will actually use.

Why does a flag say present when the value is missing?

Because preview_properties reports that a field exists in the database, not that the call returns it. money_raised was marked present on some rounds and had no value. Treat the flag as a hint about what a seat would show and never read it as data.

What does a list cost?

Each company is a search and a rounds call, $0.04 (Verified 2026-10-16), plus a profile call at $0.02 if you add it. A thousand companies is $40.00 without the profile. We ran two calls today, $0.04 in total.

Which endpoint should I use for which job?

Prices Verified 2026-10-16 with monid inspect. The first and third rows are the ones we ran. The profile endpoint we inspected and did not run, so this post does not describe its fields: spend one call on a company you know well before you build a step on it. Current prices for every endpoint are on monid.ai/tools, and other company data on the same key and balance is under the company enrichment tools.

When is the Pro seat the better purchase?

Four cases, and they are common.

You query the market. "Every Series A in this sector since January" is a database question. A lookup endpoint answers it only for companies you already named, and the seat's search is built for it.

You need amounts, dates, valuations or investors. None of the four came back on our two calls. If a number goes into a model or a memo, buy the licensed access.

A person browses. Saved searches, alerts, list building and export are what an interface is for, and a loop of calls does not replace someone's morning with the data.

Many people ask many questions. The seat's marginal cost per question is zero. Past a certain volume of human use, a fixed fee beats any per-lookup price, and your own break-even arithmetic will show where.

And the disclosure: we resell these endpoints. The sharpest thing in this post is that the calls return the shape of a funding history and none of its size, which is not a flattering description of what we sell, and the advice to skip the profile call is advice to spend less with us.

Conclusion

Pay by the seat when people browse the database and need its numbers; pay by the call when a pipeline looks up companies it already has. Through Monid a lookup is two calls at $0.02 each (Verified 2026-10-16), billed as you run them, with one key for this and every other tool and about a minute to set up.

Two things matter more than the choice. Decide first whether the fields you need are the ones that come back, because the call returned the category, size band, status and typed rounds for Stripe and not one amount. Then handle the resolve step honestly: 70 matches for one short name means you pick by rank or permalink, never by position.

How we estimated the time saved: weighing seat vendors and contracts for a lookup job is about 2 to 4 hours, wiring a resolve-then-rounds chain with id handling is 2 to 4, and the first month of upkeep for a scraper or a changing export is 2 to 6. That is 6 to 14 hours. The 1 minute is the install, key and first call once you already have a Monid key.

Free next step: run monid inspect -p crunchbase -e /get_company_funding_rounds, then search one company you know well and count the matches. Start at monid.ai.

FAQ

Published September 2026.