Attio publishes a clean four-tier ladder. Free, Plus at $44 a user a month, Pro at $99, and Enterprise on a quote. Both billing bases are printed, the annual saving is stated, and the feature comparison runs to sixty-odd rows behind no demo form at all. It is one of the most transparent pricing pages in this directory. The problem is a number it does not show.
Plus stops at ten seats. Not ten seats included, with the eleventh billed on top. Ten, full stop, after which the tier is unavailable and every existing seat reprices from $44 to $99.
What follows: the eleventh-seat arithmetic in full, where the 20% annual badge actually lands, the two credit meters running at once and what Attio's own top-up rates say the Pro allowance is worth, the limit that binds before seats do, and a correction this site had to publish about its own Attio figure.
What does Attio cost?
$44 a user a month on Plus and $99 on Pro, billed monthly. $35 and $79 billed annually. Free is genuinely free for up to three seats with no card and no expiry. Enterprise is a quote and is annual only.
| Tier | Monthly | Annual | Seat cap | Objects | Records |
|---|---|---|---|---|---|
| Free | $0 | $0 | 3 | 3 | 50,000 |
| Plus | $44/user/mo | $35/user/mo | 10 | 5 | 250,000 |
| Pro | $99/user/mo | $79/user/mo | Unlimited | 12 | 1,000,000 |
| Enterprise | Quote | Quote, annual only | Unlimited | Not published | Not published |
This site records $44, the monthly-billing rate, because it is the price a buyer can pay without committing a year. The annual rate is noted rather than recorded, which is the convention across every entity here and exists so a comparison table does not silently import a twelve-month commitment into half its rows.
The seat column is the one that decides the purchase, and it is printed on Attio's page as a feature bullet reading Up to 10 seats rather than as a boundary.
What a seat cap is, and why it is not a feature
A seat cap is a ceiling above which a tier cannot be bought at all.
That is different from a seat allowance, where the extra person is billed on top at the same rate, and the difference is the whole of this page.
An allowance charges you for one more person. A cap charges you for everyone again, at a higher rate.
The practitioner literature on this pattern is about consumer pricing rather than about Attio, and it is worth knowing it exists. A Hacker News thread titled "Companies use drip pricing to overcharge consumers" drew 317 points and 416 comments in January 2023. That is a large number of people characterising a mechanism, not a finding of fact about any vendor, and it is cited here because the mechanism is the same one: a cost that is knowable in advance and is not put where the buyer is looking.
The eleventh seat costs $649 a month
| Team size | Tier | Rate | Monthly bill | Annual-basis bill |
|---|---|---|---|---|
| 3 | Free | $0 | $0 | $0 |
| 4 | Plus | $44 | $176 [computed] | $140 [computed] |
| 9 | Plus | $44 | $396 [computed] | $315 [computed] |
| 10 | Plus | $44 | $440 [computed] | $350 [computed] |
| 11 | Pro | $99 | $1,089 [computed] | $869 [computed] |
| 12 | Pro | $99 | $1,188 [computed] | $948 [computed] |
| 20 | Pro | $99 | $1,980 [computed] | $1,580 [computed] |
| 50 | Pro | $99 | $4,950 [computed] | $3,950 [computed] |
The ratio being identical on both billing bases is the tell that this is structural rather than promotional. A 20% discount applied to both sides of a ratio leaves the ratio alone, and what survives is the shape of the plan design. A team sizing itself at eight or nine has no way to see this from the pricing page, because a cap presented as a feature bullet reads like a limit on what you get rather than a limit on what you can buy. Budgeting from $44 and landing on $99 is a 125% variance on the largest recurring line in most early-stage stacks.
None of that is a criticism of the price. Pro carries call intelligence, sequences, permission controls and twelve objects against five, and a growing team may want every one of them. The criticism is about where the number is, not what it is.
Where the 20% annual badge actually lands
Attio badges a 20% annual saving and the badge is honest.
It is also, on both tiers, slightly better than 20% once the rounding is done, which is the opposite of the usual direction.
| Tier | Monthly | 20% of monthly | Printed annual | Actual discount |
|---|---|---|---|---|
| Plus | $44 | $35.20 [computed] | $35 | 20.45% [computed] |
| Pro | $99 | $79.20 [computed] | $79 | 20.20% [computed] |
Attio rounds down to the whole dollar and the buyer keeps the difference. It is 20 cents a seat a month on Plus and 20 cents on Pro, which is trivial money and non-trivial evidence: a vendor that rounds a discount in the customer's favour is not the vendor you expect to find hiding a repricing event two bullets away. Both readings hold. The page is unusually complete and the seat cap is unusually consequential, and neither cancels the other.
| Scenario | Year, monthly billing | Year, annual billing | Saved |
|---|---|---|---|
| 1 seat, Plus | $528 [computed] | $420 [computed] | $108 [computed] |
| 10 seats, Plus | $5,280 [computed] | $4,200 [computed] | $1,080 [computed] |
| 11 seats, Pro | $13,068 [computed] | $10,428 [computed] | $2,640 [computed] |
| 20 seats, Pro | $23,760 [computed] | $18,960 [computed] | $4,800 [computed] |
Every one of those twelve figures is twelve times a published monthly rate. Attio does not print annual totals and neither should you, without saying whose arithmetic they are.
Two credit meters, running at the same time
Attio's AI features run on credits, and there are two separate allowances rather than one.
| Free | Plus | Pro | Enterprise | |
|---|---|---|---|---|
| Seat credits | 100 /user/mo | 500 /user/mo | 1,000 /user/mo | 2,500 /user/mo |
| Workspace credits | 250 /workspace/mo | 1,500 /workspace/mo | 10,000 /workspace/mo | Custom, not published |
| Plus to Pro multiple | n/a | n/a | 2x seat, 6.7x workspace [computed] | n/a |
The two pools scale at completely different rates across the same tier boundary.
Seat credits double. Workspace credits go up 6.7 times. Whatever draws on the workspace pool is what Attio expects to grow with the organisation rather than with headcount. The pricing page does not say which features draw on which pool, and that is the question to put to a salesperson before signing, because a team whose usage is workspace-shaped is buying a very different product from one whose usage is per-seat.
What the Pro premium actually buys
Attio sells credit top-ups at published rates, which is unusual and useful: it makes the vendor's own included allowances priceable in the vendor's own currency.
| Top-up | Monthly | Per credit | Annual | Per credit |
|---|---|---|---|---|
| +5,000 | $85 | $0.0170 | $70 | $0.0140 |
| +10,000 | $150 | $0.0150 | $120 | $0.0120 |
| +25,000 | $330 | $0.0132 | $260 | $0.0104 |
| +50,000 | $595 | $0.0119 | $475 | $0.0095 |
That is a genuine volume curve. The unit price falls 30% across a tenfold range and falls the whole way, which most published discount ladders in this directory do not manage. So credits are not what you are buying at Pro, and that is worth knowing before a negotiation. Roughly a tenth of the premium is allowance and the rest is features and the removal of the seat cap.
It also produces a clean decision rule. If credits are your constraint, buy credits. Fifty thousand a month costs $595, and seven Pro seats cost $693 [computed], so the largest top-up on the ladder is cheaper than seven seats of the tier you were about to upgrade to.
The limit that binds before seats do
Objects. Three on Free, five on Plus, twelve on Pro. For a CRM whose entire selling point is that you shape it to your own data model rather than conforming to somebody else's, five is tight. The buyer Attio is aimed at is precisely the buyer who wants a seventh object, and that buyer hits the wall well before the tenth hire.
Two other ceilings are worth writing down before you sign. Records go 50,000, 250,000, then a million. Emails are per month and go 200, 1,000, then unlimited, which means the entry tiers are not for sending and should not be budgeted as though they were.
Attio next to the CRMs it gets compared with
| Tool | Entry paid, monthly | Free tier | Self-serve | How it gets expensive | Price read |
|---|---|---|---|---|---|
| Attio Plus | $44/user | Yes, 3 seats | Yes | Seat cap at 10 reprices everyone | 2026-08-08 |
| Attio Pro | $99/user | n/a | Yes | Two credit pools, no published map | 2026-08-08 |
| Pipedrive | $14/user | No | Yes | Thin marketing and support features | No read date recorded |
| Close | $19/user | No | Yes | Calling metered separately on top | 2026-08-12 |
| HubSpot | $20/user list | Yes, unlimited users | Yes | Contact tiers and seat types | 2026-08-12 |
| Salesforce | $25/user | No | No | Admin capacity, not licence cost | 2026-08-12 |
| folk | $30/user | No | Yes | Four consumption meters, no published allowance for any | 2026-08-11 |
| Clay | $185 flat | Yes | Yes | Two sliders, headline is their sum | 2026-08-11 |
Attio's entry rate is the second highest on that list and its free tier is one of only three. The comparison that matters is not the rate though. It is the fifth column, because every one of these products is cheap at signup and each becomes expensive by a different mechanism.
The Pipedrive row is left honest rather than tidy. That record carries no `pricingVerifiedOn` date at all, which means it has not been re-read since the initial research pass, and a cell that says so is worth more than a cell that guesses.
folk is the row that makes Attio look generous. It names four consumption meters in one bullet, enrichments, messages, account sync and AI fields, and publishes an allowance for none of them on any tier. Attio publishes two pools with numbers in both. Undocumented is bad; unnumbered is worse.
How to check the cliff in twenty minutes
You do not need an account for most of this.
- Write down your headcount twelve months out, not today. If the number is 9, 10 or 11, stop and read the rest of this list carefully. If it is 4, the cliff is a year away and Plus is correct. Two minutes.
- Open attio.com/pricing and find the seat bullet on the Plus card. Confirm it still reads as a ceiling rather than an allowance. If Attio has changed this since 8 August 2026, everything on this page is stale and you should trust the page over us. Three minutes.
- Count the objects you need. List the things your business tracks that are not people, companies or deals. If there are more than two, Plus is already the wrong tier and the seat cap is not your first problem. Five minutes.
- On the free tier, run your AI workflow for a week and read the credit counter. Record seat credits and workspace credits separately, because the two pools scale 2x and 6.7x across the same boundary and only one of them may be your constraint. Ten minutes of setup, one week of waiting.
Pass condition: your twelve-month headcount is under nine, you need five objects or fewer, and your weekly credit burn leaves headroom in both pools. Fail condition: any one of those is false, in which case price Pro from the start and treat Plus as a trial rather than as a plan.
What breaks in month three
You hire somebody. That is the whole failure mode and it is the one nobody plans for, because hiring is a good thing happening and a 2.48x invoice does not feel like a consequence of it.
The credit pool you were not watching runs out. Two meters means two ways to run dry, and Attio does not publish which features draw on which. A team that budgeted seat credits and consumed workspace credits will find out in the middle of a month.
You need enterprise controls and cannot buy them monthly. Enterprise is annual only and unpriced, and so is the workspace credit allowance attached to it. A regulated buyer discovering this in month three has a procurement cycle in front of them, not an upgrade button. And the sixth object arrives. It always does, because the reason you chose this CRM is that your business does not fit five.
This site published $34 for Attio
The lesson generalises past Attio. A figure that is nearly right is more dangerous than one that is obviously wrong, because nothing about it triggers a re-check. Every price in this directory now carries a read date for exactly that reason, and the rows without one are marked rather than assumed current.
Which tier should you actually buy?
Under three people: Free.
It is a plan rather than a demo, with three seats, no card, no expiry, and real-time contact syncing and automatic enrichment included, which is the thing most CRMs put behind the first paywall.
Four to eight people, with five objects or fewer, and no hiring plan: Plus at $44. The cliff is more than a year away and you will have re-priced the stack by then anyway.
Nine or more, or hiring: Pro at $99, from day one. Treat Plus as a trial rather than as a plan. Budgeting from $44 and landing on $99 is a 125% variance on the largest recurring line in most early-stage stacks, and it arrives on the day you do something good.
Credits are the only constraint: stay where you are and buy credits. $595 a month for fifty thousand is less than seven Pro seats at $693 [computed], and it is the one lever on this pricing page that moves in your favour.
So what should you budget?
Budget the tier above the one you want, and budget it annually. On Attio specifically that means $79 a seat rather than $44, because the difference between those two numbers is a hiring decision rather than a purchasing one, and hiring decisions are made by people who are not looking at the CRM invoice. Eleven seats on Pro is $10,428 a year on the annual basis and $13,068 on the monthly one [both computed]. Ten seats on Plus is $4,200 and $5,280. The honest planning number for a team that intends to grow is the first pair, not the second.
If you are already over ten, none of this applies and Pro is simply the price. At $99 against Salesforce's $25 plus a dedicated administrator, it is not the comparison it looks like.