How-to · 3 steps
How to research a company
with Monid.
Give this to your agent
$set up https://monid.ai/SKILL.md, then brief me on <company>.com and show the cost of each step
and let it take it from there.
#JobWhat came backCost
1The brief2,600 staff, industry, HQ, market cap$0.05 ✓
2Departments965 engineering, 372 sales$0.00 ✓
3Their stack240+ tools, same call$0.00 ✓
4The tech detectorfive shallow signals$0.0356 ✓
5Funding7 rounds, 23 investors, $432M$0.07 ✓
6News10 articles in 30 days$0.0009 ✓
7Lookalikes663 matches, noise included$0.05 ✓
8The degraded one502 twice, billed nothing$0.00 ✓
9Totalnine runs, seven billed$0.2386 ✓
one company · six questions · three answered by one callagent running$0.2386
Real run, 2026-09-24. A public company, public figures.
Step 1
Set up Monid.
One line. It installs the CLI and asks for an API key from app.monid.ai. New accounts start with $1.00.
Say this to your agent
>set up https://monid.ai/SKILL.md
Step 2
Pick what you need. Say it in one sentence.
Click a job. Paste the sentence, fill in the brackets.
Brief me on this company
Say this to your agent
>enrich <company>.com with apollo on monid and give me the industry, headcount, location and description. read the whole response before you call anything else: it carries more than the brief
What the agent runsdone
apollo /organizations/enrich$0.05
the-companies-api /enrich_company_from_domain · not run$0.01
context.dev /brand/retrieve · not run$0.009
clay /enrichment/company-employee-count · not run$0.10
pdl /v5/company/enrich · not run$0.10
akta /v1/company/enrichment · not run$0.125
What came backone public company · 2026-09-24
What one $0.05 call returnedjobs answered
the brief2,600 staff, the industry, a San Francisco HQ, founded 2014, market cap $8.3B
and morethe same payload carries the departmental headcount and a 240-item technology list
what that meansone call answers three of the six jobs on this page
the disagreementthe structured headcount says 2,600 while the prose in the same response says about 2,375, and on a second company a round's own description said total funding reached $130M while the rows beside it summed to $175.1M
what the description promises and omitsthe endpoint's own summary says it returns funding history and the payload carries no funding field at all; a second company's response was 53 KB with none
$0.05per company
Who works there, by department
Say this to your agent
>from the enrichment response for <company>.com, give me the departmental headcount sorted by size. do not make a second call: it is already in what you paid for
What the agent runsdone
apollo /organizations/enrich$0.05
hunterio /discover/people · not run$0.10
clay /enrichment/company-employee-count · not run$0.10
ploid /linkedin/company · not run$0.01
What came backone public company · 2026-09-24
Headcount by departmentpeople
the shape965 in engineering against 372 in sales, so this is a product company, not a sales-led one
the tailsupport 258, HR 196, finance 161, and a dozen smaller functions
what it costnothing, because it arrived inside the enrichment call
the habita second endpoint for this would have been $0.10 and told you less
$0.00already in the brief
What technology they use
Say this to your agent
>list the technologies <company>.com uses from the enrichment response. do not add a dedicated detector on top: it is a second charge for a shorter list, and on our runs it returned five signals against three hundred
What the agent runsdone
apollo /organizations/enrich$0.05
api.strale.io /x402/tech-stack-detect$0.0356
context.dev /brand/retrieve · not run$0.009
What came backone public company · 2026-09-24
the deep listmore than 240 tools, from cloud infrastructure to HR, sales and security systems
the detector$0.0356 for five signals it read off the public HTML: a CDN, analytics, a framework and two marketing tags
why the gapone reads a company record, the other reads one page's source
it is not dearer, it is redundantthe detector is $0.0356 against the enrichment call's $0.05, so it is cheaper per call; the point is that you already paid for a list sixty times longer, so running it is a second charge for a worse answer
and it can be wrongon a marketing-automation company the detector reported no marketing automation, because it is reading one page's headers rather than a company record
$0.00already in the brief
Funding history and investors
Say this to your agent
>get the funding history for <company>.com with fundable on monid using company/search then company: two calls, a cent and six cents. do not reach for the endpoint that takes the domain in one call, it is priced per result with a page size of ten and bills six times as much for the same rounds
What the agent runsdone
fundable /company$0.06
fundable /company/search$0.01
crunchbase /get_company_funding_rounds$0.02
crunchbase /search_organizations$0.02
surf /fund/portfolio · not run$0.006
What came backone public company · 2026-09-24
the history7 rounds and 23 investors, $432M raised in total
the latesta dated $195M secondary sale at a $6B valuation, with the buyers named
the failurethe other provider's funding endpoint 502'd twice and billed nothing; the catalog had already flagged it degraded
the shortcut that costs 6xa sibling takes the domain directly with no lookup, which looks like the cheap route; it is priced per result with a silent page size of ten and billed $0.42 for the same seven rounds
and then the investorsthat sibling returns investors as bare ids, so resolving the names the sentence asks for is another call per round; the honest budget on that route is eight calls and $0.84 against $0.07 on this one
three of five are per resulton this provider only the lookup and the company call are per call; the deals, companies and rounds endpoints all multiply by the page size you did not set
$0.07lookup plus history
What happened in the last 30 days
Say this to your agent
>search news for <company> in the last 30 days with context.dev on monid, passing the dates as epoch milliseconds rather than as date strings, and group the rows by story_id before you count them: ten rows is usually fewer stories, and most of them are press releases
What the agent runsdone
context.dev /news/search$0.00009 / result
surf /project/ai-news · not run$0.012
akta /v1/news · not run$0.0005 / result
apollo /news_articles/search · not run$0.05
What came backone public company · 2026-09-24
the setten real articles across a 30-day window, with more available behind a flag
the one that mattersa maximum-severity vulnerability covered by two outlets, next to a routine product release
the price$0.0009 for the whole set, the cheapest job on this page by a wide margin
the duplicatethe same press release came back twice under two tracking urls, and both count as results; on a second company ten rows were five distinct stories and eight of the ten were press releases
the date formatthis endpoint wants epoch milliseconds, so a plain date string is rejected before the call is made, which costs nothing but stops the job
and the filter leakstwo rows came back with a null published date despite a published-after filter being set
$0.0009ten results
Companies that look like them
Say this to your agent
>find companies like <company> with apollo on monid using its industry keywords and headcount range, then read the breadcrumbs field back to me. the keyword filter is an OR, the subject appears in its own results, and the rows carry no industry or headcount, so tell me what you cannot filter rather than pretending to clean the list
What the agent runsdone
apollo /mixed_companies/search$0.05
blockrun.ai /api/v1/exa/find-similar$0.0121
the-companies-api /search_industries · not run$0.01
pdl /v5/company/search · not run$0.10
What came backone public company · 2026-09-24
the set663 matches, with real peers on the first page
the noisethe subject came back inside its own lookalike list; on a second company it was the fourth result, beside five staffing and jobs firms and a car maker, an airport and a newspaper, with three of twenty-five usable
the wrong toolthe semantic-similarity endpoint billed $0.0121 and returned ten more pages of the subject's own website
the cause is in the responsethe breadcrumbs field says the keyword filter is an ANY match, so one shared word is enough to pull a company in, and nothing in the request suggests that
you cannot clean it from these rowsthe results carry no industry, no headcount and no location, so filtering out staffing firms means enriching every row at $0.05 each, which is more than the search cost
the lessonsimilar pages and similar companies are different questions; the description was accurate, the fit was not
$0.05per search, 663 matches
Step 3
Take the cheapest route with Monid.
Read the whole response before you buy another one. One enrichment call answered three of these six jobs.
News · $0.00009 a result
News search$0.0009ten articles
Brand retrieve$0.009name, logo, socials
The brief · cheapest first
Companies API$0.01
Apollo enrich$0.05
Fundable company$0.06
Clay headcount$0.1
PDL company$0.1
Fill the gaps
Funding history$0.01
Tech detector$0.0356only if the stack list is thin
Back to you
rowbriefed bycost
1Fundable company$0.1309
2nothing moved$0.0009
3Companies API$0.0209
4Fundable company$0.1309
5PDL company$0.2309
5 of 5 rows$0.5145
Say this to your agent
>brief me on these companies, cheapest source first, and show the cost per company: <paste domains>
“read the whole response”one enrichment call carried the brief, 19 departments and over 300 technologies
“quote the structured field”two responses disagreed with their own prose on headcount and total raised
“exclude the subject”the lookalike search returned the company itself and three staffing agencies
“check how it is priced”the one-call funding shortcut is per result with a page size of ten, and billed six times the two-call route
One key. 1,700+ tools.
Company data is the biggest aisle in the catalog.
One call, most of the briefapollo · $0.05 / call
News and brand datacontext.dev · from $0.00009
Enrich from a domainthe companies api · $0.01 / call
Rounds and investorscrunchbase · $0.01 to $0.02
Deals and valuationsfundable · $0.01 to $0.06
Headcount and domainsclay · to $0.10
Who works therehunterio · to $0.10
The company on LinkedInploid · $0.01 / call
Deep company recordspdl · $0.10 / call
Enrichment and newsakta · to $0.125
Funds and eventssurf · $0.006 to $0.012
Accelerator companiesycombinator · $0.01 / call
Company searchcontactout · $0.02 / call
Tech stack and checksstrale · from $0.02376
+ 1,700 more across 55 providers
Browse the catalog →Read more
View all →
Apollo, ZoomInfo and the alternatives
What the big databases charge for the same rows, and what an agent pays instead.
Cheapest source first, then fall through
Once the brief names a person, finding their address is the next call.
Extract page content for RAG
When the answer is only on their site, reading the page is the next call.