The B2B Data Free Tier Census: What 37 GTM Tools Actually Give You

Cleanlist fetched the live pricing and help pages of 37 B2B data and GTM vendors on September 1, 2026 and recorded, in identical columns, what each one gives away free. Not one of the 37 pages states that a credit card is required: 14 say explicitly that no card is needed, 18 say nothing at all, and 5 have no trial for the question to apply to. Fifteen vendors run a recurring free plan and 9 offer a time-boxed trial. Only 7 publish that bulk CSV upload works on the free tier, and 2 of those are hard-capped. Among the 9 vendors that publish a convertible monthly allowance, the median free plan is worth 40 verified emails a month, and across the 7 that publish a monthly phone allowance the median is 4 phone numbers. The full 37-row dataset is below under CC BY 4.0.

Last updated: September 1, 2026. All 37 pages were fetched on September 1, 2026.

Which of these tools have a real free plan, and which only have a trial?

Of the 37 vendors, 15 run a recurring free plan and 9 offer a time-boxed trial. Those two columns are independent rather than a partition, so read them as overlapping sets rather than as shares of 37. Apollo, Clay and Cleanlist do both, which makes 21 distinct vendors offering one free entry or the other. The remaining 16 offer neither: 8 give a one-time grant that never refills, 4 published nothing our collector could read on either column, and 4 (Cognism, HubSpot Breeze Intelligence, Instantly and NeverBounce) reference a free entry without a countable allowance. The split is sharply categorical. Five of the 11 prospecting databases carry a recurring free plan: Apollo, Cleanlist, LeadIQ, Lusha and Seamless.AI. Two of the two browser-extension tools do, Kaspr and Surfe. But zero of the four sequencer-shaped products do: Instantly, Smartlead, Lemlist and Amplemarket all publish paid plans only, with Lemlist offering 14 days of full Multichannel access and no card. Snov.io is the outlier in that group, running a Trial plan that renews every 30 days indefinitely. Cognism publishes no self-serve free entry of any kind, only a gated data sample and a sales call.

Do any of these tools require a credit card to try them?

No. Across all 37 pricing pages fetched on September 1, 2026, the value yes appears zero times in the credit-card column. Fourteen vendors state plainly that no card is required, and they are the complete list: Bouncer, Cleanlist, Clearout, Datagma, Findymail, FullEnrich, Instantly, Kaspr, Lemlist, NeverBounce, Smartlead, Surfe, ZeroBounce and ZoomInfo. Eighteen say nothing at all, which includes Apollo, Clay, UpLead, Amplemarket, LeadMagic, Snov.io, Emailable and Kickbox. Five have no card question to answer because they have no trial. Nobody in this set demands a card at the door. What 18 of them do instead is leave the question unanswered, which is the finding worth taking away. Silence on the one term a buyer checks first is what sends people searching for "free trial no credit card" rather than simply signing up.

How many verified emails does a free plan actually give you per month?

Nine vendors publish both a recurring monthly allowance and a per-email credit price, which is what it takes to answer this question from a pricing page alone. Among those nine the median is 40 verified emails a month and the mean is 38.9. Datagma leads at 90. Persana, Hunter and Snov.io sit at 50, though Hunter and Snov.io both draw finding and verifying from the same pool, so 50 emails means no verifications left. Lusha gives 40. Cleanlist gives 30. Surfe and Kaspr give 20. ZeroBounce, which is a verification service rather than a finder, gives 10 Email Finder credits alongside its 100 monthly validations, and its own pricing page states that "ZeroBounce Email Finder uses 20 credits per successful query". Run through this census's normalisation rule, those 10 credits buy 0.5 verified emails a month, the smallest of the nine monthly allowances and the reason the mean now sits below the median. Eighteen of the 37 vendors publish enough to make this conversion at all, 13 publish an allowance without a convertible unit price or publish nothing, and 6 are verification-only tools where the question does not apply.

How many phone numbers do you get free?

Almost none, and this is the sharpest finding in the census. Seven vendors publish a recurring monthly free phone allowance. Added together, all seven give a combined 25 phone numbers a month. Persana, Surfe and Kaspr give 5 each. Lusha gives 4. Cleanlist and Datagma give 3 each. Clay gives zero, because its own comparison table states that the Free plan does not include phone number enrichment. Apollo gives phone access "for a limited time" and then, in its own words, "phone number access will be disabled". Direct dials are the most expensive record type in this market, which is exactly why nobody hands them out, and it means no free tier in this set can answer the question "is the phone data any good". The Cleanlist 500-Lead Enrichment Benchmark needed 500 stratified leads to measure 85% direct dial coverage. Cleanlist's own Free plan buys 3 of those 500 numbers a month.

Can you upload a CSV on the free plan?

This is the motion that proves an enrichment product, and it is the one most often withheld. Only 7 of 37 vendors publish that bulk CSV upload works free, and 2 of those are hard-capped: Clay allows CSV into tables but caps the Free plan at 200 rows per table and excludes bulk enrichment above 50,000 records, and Kaspr allows 3 CSV files a month at 25 rows per file. The clean yeses are Datagma at up to 50,000 rows, FullEnrich, and the verification services Emailable, NeverBounce and Verifalia. Six vendors publish an outright no, and Cleanlist is one of them: CSV upload is a Starter feature, available during the 14-day Scale trial and not on the $0 Free plan. LeadIQ puts file enrichment on Enterprise only. The remaining 24 vendors say nothing.

Can you get your data out?

Export is the second gate, and it is even less legible than the first. Twenty-six of the 37 vendors say nothing about export terms on the free tier. Seven publish that export works free: Apollo, Clay, Cleanlist, Kaspr, LeadIQ, Seamless.AI and UpLead. Three publish an outright no, and they are worth naming because the effect is severe. That is 36. The 37th is HubSpot Breeze Intelligence, whose column reads n/a because a free HubSpot seat cannot spend a credit at all, so there is no export question underneath. Snov.io's Trial plan gives you 50 credits every 30 days and states that export is not available in Trial, so the data stays inside the product. Skrapp gives 50 credits and puts List Exports on Professional and above. Cognism gives nothing self-serve at all. Four vendors meter export separately from finding, so an export is a second charge on a record you already paid to reveal: Apollo, Clay, Kaspr and UpLead. Cleanlist does not meter export, and search costs 0 credits on every plan including Free.

How many seats does a free plan give you?

Only 15 of 37 vendors put any value at all in the seat column, and one of those is Cognism, which has no free tier for a seat to sit on. A buyer evaluating for a team cannot answer this from the pricing page 22 times out of 37. Where it is published, the answer is usually one. Eight vendors give exactly one seat: Apollo after its initial access period, Cleanlist, Kaspr, LeadIQ, Seamless.AI, Skrapp, Surfe and UpLead. Four give unlimited seats on the free tier, which is unusually generous and worth crediting by name: Clay, FullEnrich, Hunter and Persana. Datagma publishes 10 seats included, on the free plan and every paid plan alike. HubSpot's Breeze Intelligence gives zero credit-eligible seats, because its documentation states that free users and view-only seats cannot spend HubSpot Credits at all.

What expires, and when?

Three different clocks run in this set and vendors rarely label which one they are using. Eleven vendors reset a monthly allowance. Eleven grant a one-time bucket with no refill, including FullEnrich, BetterContact, Dropcontact, Findymail, Emailable, Bouncer, Clearout and Kickbox. Verifalia runs the only daily clock, 25 credits resetting at midnight GMT with no accumulation, so an unused day is simply gone. Several verification vendors go the other way and promise credits that never expire: ZeroBounce, Emailable, Bouncer and Clearout all state it. NeverBounce is the exception, expiring purchased credits 12 months after purchase. Apollo expires a capability rather than a balance, switching phone access off after an initial period. Cleanlist expires the 14-day Scale trial, which drops to the $0 Free plan rather than to a locked account.

What is the hardest cap most people hit first?

Rarely the credit balance. On Clay it is 200 rows per table, which stops a realistic list test long before 100 data credits run out. On Kaspr it is 25 rows per CSV file, three files a month. On Skrapp it is the absence of list exports. On Snov.io it is the absence of export, bulk search and bulk verification together. On HubSpot Breeze it is that enrichment requires a paid seat, so the free CRM enriches nothing. On Emailable it is that the smallest paid purchase after the 250 free credits is 5,000 credits. On Cleanlist it is CSV upload sitting on Starter, which is why the 14-day Scale trial exists and why the honest advice is to run the bulk test during the trial rather than after it. On UpLead it is 5 contacts in seven days.

What do the email verification tools give away?

Verification is the most generous corner of this market relative to the size of the job, because a verification costs the vendor a fraction of what a phone number costs. ZeroBounce runs a Freemium plan at $0 a month with 100 validation credits that refill every month, plus 10 Email Finder credits, one inbox test, one server test and one blacklist monitor, gated on signing up with a business or premium domain. Those 10 finder credits are worth half a lookup, because the same page prices the Email Finder at 20 credits per successful query. Verifalia gives 25 credits a day, which is roughly 750 a month if you use every day, and they do not bank. Emailable gives 250 credits once. Bouncer, Clearout and Kickbox each give 100 once, and Bouncer and Clearout state that those credits never expire. NeverBounce advertises "free test credits" and never states the number.

Which vendors give you effectively nothing?

Two, for different reasons. Cognism publishes no free plan, no trial and no price, offering a free data sample on request and a sales call, so there is no way to touch the product without a conversation. HubSpot's Breeze Intelligence, the successor to Clearbit, is stricter still: its own credit documentation states that only users with paid seats can spend HubSpot Credits and that free users and view-only seat users cannot, so a free HubSpot CRM enriches exactly zero records. Amplemarket, Instantly, LeadMagic and Smartlead all reference a free trial without publishing an allowance, a length, or both, which lands in the same practical place for a buyer trying to plan an evaluation on a Tuesday afternoon. Naming this is not a criticism of the products, which several buyers rate highly. It is a criticism of the pages.

Why could you not read five of these pricing pages?

Five of the 37 pages did not deliver their free-tier figures to a plain HTTPS fetch on September 1, 2026, and the census records that rather than filling the gap from someone else's blog. Four of them never served a plan table at all: RocketReach returned an 11 KB shell, ContactOut loads its plan values from an API after render, Prospeo renders its table client-side, and Ocean.io returned HTTP 403 on every attempt. Prospeo needs one extra sentence, because our B2B Data and GTM Pricing Index, published the same day, quotes Prospeo's plans in full. That index drives a headless Chromium browser and executes the page's JavaScript; this census runs a plain HTTPS fetch and executes none. We re-fetched prospeo.io/pricing over plain HTTPS on September 1, 2026 to check, and it still served no plan table. Both pages are right about their own collector, and the disagreement is a fact about how the page is built rather than about what Prospeo charges. Apollo is a different case worth stating precisely. Its page loaded in full, and its own FAQ text reads "{creditsNumber} credits per month to get started" with the number left as an unfilled template variable. Two more pages served partial answers: ZoomInfo describes a trial with "usage limits" it never quantifies, and NeverBounce offers free test credits without a count.

What does Cleanlist give away, in the same columns?

Every new Cleanlist workspace starts on Scale for 14 days: 250 credits, 3 seats, no card, with every feature except the public API and MCP server. When the trial ends the account drops to a $0 Free plan with 30 credits a month and 1 seat, which is 30 verified emails or 3 phone numbers at 1 credit per email and 10 per phone. Search costs 0 credits on every plan including Free, CSV export is included on Free with 100 leads per list, and there is no separate export meter. CSV upload is not on Free. Credits are charged only when data comes back, so an empty lookup costs nothing. Paid plans are Starter $79, Pro $229 and Scale $599 a month, with annual billing 25% off.

Which free tiers can actually support a real evaluation?

Match the tier to the question you are asking, because the free tiers in this set support four different shapes of test and no tier supports all four. To answer "does this find the person I am looking at right now", any recurring free plan works, and the extension tools are built for exactly that: Kaspr, Surfe and Lusha. To answer "how good is the data on a list of 50 people whose answers I already know", you need bulk upload plus export together, which narrows the field to Datagma, FullEnrich, Clay within its 200-row cap, and Kaspr within its 25-row cap. To answer "can three people use this for a week", you need seats, which means Clay, FullEnrich, Hunter, Persana or Datagma. To answer "is my existing list clean", ZeroBounce and Verifalia both refill.

Which free tiers are too small to prove anything?

Four, on the numbers rather than on opinion. UpLead's trial is 5 credits over 7 days, which is five contacts. Findymail gives 10 credits. Skrapp gives 50 credits and no way to export a list. Snov.io gives 50 credits every 30 days with export, bulk search and bulk verification all excluded, which means you can see data you cannot use. None of these are bad products and several are Cleanlist suppliers. The point is narrower: a sample of 5 to 50 records, with no bulk path and often no export, cannot separate a strong match rate from a mediocre one at any useful confidence, so a buyer who evaluates on these tiers is choosing a vendor on interface impressions. The honest read is that the free tier is a demo, and the trial is the test.

What counted as "not published"?

A cell reads not published when the vendor's own live page, or the help page it links to, did not state the value in text our collector received on September 1, 2026. It does not mean the vendor hides the number, and it does not mean the number does not exist inside the product. Three distinct situations collapse into that one label, which is a real weakness of this dataset: the vendor genuinely does not state it anywhere, the vendor states it behind JavaScript our fetch did not execute, or the vendor states it on a page we did not think to follow. The figure_served_by_vendor_page column separates the second case out for the 7 vendors where it applies. Anywhere the label is wrong, the fix is a corrected page rather than a defended one.

What are the limits of this census?

One snapshot on one day, September 1, 2026, from one country, over plain HTTPS with no JavaScript execution, which is the single largest weakness here and the reason 5 pages came back empty. Thirty-seven vendors chosen by a competitor of several of them, unweighted by market share, so this is a shortlist of what a sub-300-employee B2B SaaS buyer sees rather than a survey of the category. Nothing was verified by creating an account, so every figure is what the vendor advertises rather than what the product delivers. Pricing pages change without notice and several in this set run limited-time offers, including a 30-day Lemlist trial promotion and a 30%-off Kickbox code visible on the day. And Cleanlist ran the census, is in it, and loses the bulk-upload column.

What did we change after the first internal review?

An adversarial recheck ran over this page and its CSV before publication, with a mandate to break both. Four things moved.

ZeroBounce's row was wrong by 20x. We first published 10 verified emails a month for ZeroBounce, having read its 10 free Email Finder credits as 10 finds. ZeroBounce's own pricing page states that "ZeroBounce Email Finder uses 20 credits per successful query", so the correct value under this census's normalisation rule is 0.5. We re-fetched zerobounce.net/pricing on September 1, 2026 to confirm it. The mean of the nine convertible free plans falls from 40.0 to 38.9. The median stays 40, Cleanlist stays 6th of 9, and the headline is unaffected. ZeroBounce is one of our own verification suppliers, which is exactly why getting its row wrong was the worst single error in the file.

The export accounting dropped a row. 27 plus 7 plus 3 came to 37, not 38, because HubSpot Breeze Intelligence carries n/a rather than a yes or a no. The section now accounts for every row.

The free-plan sentence was written as a partition of 38 and is not one. 16 recurring plans plus 9 trials plus 4 unreadable pages summed to 29, because recurring_free_plan and time_boxed_trial are independent columns: three vendors sit in both and sixteen sit in neither. The sentence now says so and accounts for all 37 rows.

The vendor count is 37, not 38. One vendor was withdrawn for the reason given in the methodology box, and every count, list, median and category split on this page was recomputed on the 37 published rows. The findings that carry the page are unchanged: no vendor requires a card, only 7 publish free bulk CSV upload, the median convertible free plan is 40 verified emails a month, and the 7 monthly phone allowances add up to 25 numbers between them.

Where can I download the dataset?

B2B Data Free Tier Census, September 2026 (37 rows, CSV), one row per vendor, published under CC BY 4.0. Use it, republish it, build on it, credit Cleanlist and link back. Every table and every count on this page recomputes from that file alone.

Column dictionary:

How do I try Cleanlist, in the same terms as everything above?

Every new workspace starts on Scale for 14 days: 250 credits, 3 seats, no card, and every feature except the public API and MCP server. That trial is where the bulk test belongs, because CSV upload is a Starter feature and is open for those 14 days. After the trial the Free plan is $0 for 30 credits a month, so nothing is deleted and no card is ever required to keep the account. People Search and Company Search cost 0 credits on every plan, Free included, with no export meter on top. Credits are charged only when data comes back, so a lookup that returns nothing costs nothing. Enrichment runs across 25+ providers.

How should I cite this census?

Cleanlist. "The B2B Data Free Tier Census: What 37 GTM Tools Actually Give You." September 1, 2026. Data collected September 1, 2026 by direct HTTPS fetch of each vendor's own live pricing, help or terms page from a United States IP, desktop user agent, no JavaScript execution. Dataset at https://www.cleanlist.ai/data/b2b-data-free-tier-census-2026-09.csv, CC BY 4.0. If you republish a figure from this page, please link to the dataset rather than to the table, so your readers can recount it themselves. If you work at one of the 37 vendors and a row is wrong, or your page publishes something our collector missed, tell us and we will correct the row and the page on the same day.