Airtable sells you AI capacity for less than it charges for the same AI capacity. A Team seat costs $20 per user per month billed annually and includes 15,000 AI credits per billable collaborator. Buy those 15,000 credits on their own and Airtable's published top-up rate is $20 per 10,000 credits a month — $30 for the same allowance that arrives bundled in a $20 seat. That gap is a deliberate land-grab, and its architect said so out loud.
On 4 August 2026 that gap changed owner. Bending Spoons signed a definitive agreement to acquire Airtable in an all-cash deal at a $1.285 billion enterprise value, roughly $2.25 billion of equity value, closing "later in 2026" subject to regulatory approval. Bending Spoons is the company that capped Evernote's free tier at 50 notes and moved to cut three-quarters of WeTransfer's staff. The deal has not closed. That is the only good news in this story, and it expires.
What Bending Spoons Actually Bought
Bending Spoons bought a growing enterprise business, not a distressed consumer app. Airtable was running approximately $480 million in annual recurring revenue as of June 2026, growing over 20% year on year, across more than 500,000 organizations including 80% of the Fortune 100. That is a materially healthier asset than anything else in the portfolio at the point of purchase.
The price tells you why the deal happened. The release puts enterprise value at $1.285 billion and equity value at roughly $2.25 billion, the difference being Airtable's own net cash and equivalents — close to a billion dollars of the purchase price is money already sitting on Airtable's balance sheet. Against a peak of over $11 billion in 2021 and secondary-market trades around $4 billion in early 2026 on more than $1.4 billion raised, the operating business changed hands for a little over 2.5x ARR.
You do not get that multiple on a 20%-growth SaaS business without a discount for something — though none of the deal coverage says what, and a 2021 vintage reset, broad SaaS multiple compression and the softness of a secondary-market mark all sit in that gap. The component that matters to you is not whichever one explains the price. It is the one a new owner can act on unilaterally, and it is legible in Airtable's own price list: the cost of the growth.
The Line Item That Reprices First
The AI credit bundle is the single largest unpriced entitlement in your Airtable contract, and it is the cheapest thing for a new owner to change. Airtable publishes the allowances: 15,000 credits per billable collaborator on Team, 20,000 per paid user on self-serve and sales-led Business, 25,000 per paid user on Enterprise Scale at list price, pooled across the workspace or organization and reset every cycle. It publishes the standalone rate in the same document: 10,000 credits for $20 a month, scaling linearly to 400,000 for $800.
Run the arithmetic at Airtable's own list rate. The 15,000 credits inside a $20 Team seat are worth $30 standalone — 150% of the seat price. The 20,000 credits inside a $45 Business seat are worth $40, or 89% of it. Buy the discounted annual packs instead and Team still lands at $25 of credits inside a $20 seat.
Steel-man the other reading first: a top-up price is a retail price with margin baked in, not marginal cost, so this is not proof Airtable loses money on a seat. It is proof of something narrower and more useful — that the same entitlement carries two prices, and the bundled one is far lower. That spread is exactly the arbitrage a margin-focused owner closes.
Airtable's founder never hid the strategy. Discussing Superagent's pricing in January, Howie Liu told TechCrunch "we're not trying to optimize for profit margin right now" — though the rates were still being finalized at that point, and the $20-to-$200 seat range reported alongside the quote was TechCrunch's read of where the market was heading, not Liu's announcement. The June 2025 relaunch is where the strategy is unambiguous, in Liu's own words: "every Airtable plan, including our free plan, now comes with all of these AI capabilities and a generous bundling of AI credits" — a move he described in the same letter as "disrupting our own pricing model because we believe this should be the default in our new era of software."
Then work out what the allowance actually buys. Airtable's published examples price a question-and-answer action at 10 credits and analysis of a 10-page contract at 200. A Team seat's 15,000 credits is about 75 contract-scale document runs a month. If your ops team is putting real agent volume through Superagent or Omni, you are not near the ceiling by accident — you are near it because the ceiling was set generously on purpose by a management team that had explicitly stopped optimizing margin.
A new owner optimizing margin has three moves available without touching your seat price: cut the included allowance, raise the overage rate, or meter something that is free today. Any of the three lands as a bill increase you did not negotiate.
What the Playbook Has Done Three Times
Bending Spoons' post-acquisition pattern is documented, fast, and consistent across three brands. This is not speculation about intent; it is the public record.
Evernote. Bending Spoons laid off 129 Evernote employees on 17 February 2023, weeks after the deal closed, then restricted all new and existing free users to 50 notes and one notebook from 4 December 2023 — no grandfathering — with a spokesperson saying the company "has been unprofitable for years and the situation was unsustainable in the long term." Evernote's own pricing announcement set the Personal plan at $129.99 a year and gave existing subscribers 28 days' notice before their renewal date, calling it the first increase in almost seven years.
WeTransfer. Acquired in July 2024. By 8 September, Bending Spoons had confirmed plans to cut roughly 75% of a workforce of over 350. CEO Luca Ferrari's framing in that reporting is the sentence every Airtable customer should read twice: "Once the vision is clear, we try to close the gap between the status quo and the vision as quickly and as fully as we can."
Vimeo. The $1.38 billion all-cash acquisition was announced on 10 September 2025 and closed on 24 November 2025. Layoffs were reported on 22 January 2026, roughly two months later, with the company declining to say how many people were affected.
There is a data-terms precedent too. TNW's coverage of the Airtable deal notes that mid-2025 WeTransfer added terms users read as permitting AI training on uploaded files, and reversed them within days after backlash. Contract language moves in this portfolio, sometimes faster than customers can react.
The financial pressure is also new. Bending Spoons priced its Nasdaq IPO at up to $1.62 billion against a roughly $19 billion valuation and listed under BSP in July 2026. Airtable is its first deal as a public company, and it came a week after Bending Spoons brought its new and expanded credit facilities agreed since the start of the second quarter to €1.49 billion, with the term loan A and revolving commitments "available for general corporate purposes and acquisitions." Quarterly earnings calls and debt service both reward the same thing: margin.
The Steel-Man: Airtable Is Not Evernote
The strongest case against this whole article is that enterprise software does not reprice like consumer subscriptions, and Airtable is enterprise software. Ferrari has said publicly that Bending Spoons is committed to investing in Airtable for the long run, and Liu framed the deal as resources to "build the AI-native platform of the future."
Take that seriously, because parts of it are structurally true. Evernote, WeTransfer and Vimeo were consumer and prosumer products with low-friction churn and no negotiated contracts — exactly the profile where unilateral repricing works. Airtable counts 80% of the Fortune 100 as customers, and its large accounts sit on multi-year, sales-led agreements with negotiated terms, security reviews and renewal dates. Gutting a 20%-growth enterprise business is a worse trade than growing it. A buyer paying 2.5x ARR for that asset has every reason to keep the logos.
But notice what the steel-man protects and what it does not. It protects your negotiated terms. If you are on a sales-led Enterprise Scale agreement with a fixed price and a fixed term, your seat price is contractually safe until renewal. The credit allowance usually is not — it is a plan attribute referenced by the ordering document, not a negotiated number in it. And if you are one of the hundreds of thousands of organizations on self-serve Team or Business, you have no negotiated terms at all.
Your Contract Already Says They Can Do This
Airtable's standard Terms of Service grant the company every right it needs to hand you to a new owner and change your pricing, and grant you almost none in return. Read three clauses.
Clause 22.1 on assignment: the terms "may not be transferred or assigned by you without our prior express written consent, but may be assigned by us without restriction." That is a one-way assignment right. There is no change-of-control consent, no notice obligation to you, and no termination trigger when the company is sold.
Clause 6.1 on pricing: "Any change to a Subscription Plan's pricing or payment terms will become effective in the billing cycle following notice of such change to you as provided in these Terms." On a monthly self-serve plan, that is approximately 30 days between notice and a new rate card.
Clause 2.7 on service changes: Airtable "may change our Services, stop providing our Services or features of our Services... change or stop providing a particular Subscription Plan or features thereof, or create usage limits for our Services." Read that last phrase against a bundled credit allowance. Creating a usage limit is not a breach. It is a listed right.
This is the same structural exposure that turned a $39 GitHub Copilot seat into a variable bill, and the same one every recent AI acquisition has surfaced — d-Matrix buying Wallaroo, Tricentis buying Tabnine, Okta buying Permiso. The pattern does not change: the terms you never negotiated are the terms that get exercised.
What to Do Before the Deal Closes
The pre-close window is the only period where Airtable's sales organization still has its own quota, its own approval chain, and no new owner's margin targets. Use it.
This Week:
- Pull your actual agreement and find three things — the assignment clause, the price-change notice period, and whether your credit allowance is a negotiated number in the ordering document or a reference to a published plan. If it is a reference, it is not protected. If you are on the click-through Terms of Service, you have neither assignment consent nor more than one billing cycle of notice.
- Export 90 days of credit consumption per workspace from your admin billing view and compute credits per seat per month. Multiply your monthly burn by $0.002 — the published top-up rate. That number is your downside if the bundle disappears and everything reprices at list.
- List every business process with an Airtable agent or automation in its critical path — approvals, intake, customer-facing reporting, anything with an SLA attached. That list, not your record count, is your real switching cost.
This Month:
- Ask for a credit-rate lock and price protection in writing, as a contract amendment. Not an email from your account executive, which does not survive a change of control. Specify the included allowance per seat as a number, the maximum overage rate, and the term over which both hold.
- Add a change-of-control clause, or at minimum a termination-for-convenience right with a pro-rata refund triggered by a change of control. Today Airtable can assign without restriction and you cannot exit. Both halves of that are negotiable pre-close in a way they will not be after.
- Get an export commitment that covers more than records. A base exports to CSV; your automations, interfaces, permission model and agent definitions do not. Ask for a documented, machine-readable export of workflow configuration, with a defined post-termination retention and retrieval window.
- Time one exit. Rebuild a single non-critical automation on Zapier, n8n or Make and record how long it took. An untested exit plan is not leverage; a stopwatch reading is.
Before the Deal Closes:
- Decide the number now. Agree with finance what credit rate or allowance cut makes Airtable uneconomic for your workload, write it into the renewal decision, and set a calendar reminder for the close announcement. Deciding your walk-away price while you still have one is the whole exercise.
The Bottom Line
Every enterprise software cycle ends the same way: the growth-stage vendor that subsidized adoption gets sold to someone who did not sign up for the subsidy. It happened to on-premise infrastructure under private equity in the 2010s, it is happening now to the per-seat SaaS estate as agents rewrite the unit economics, and it is happening to AI capacity faster than either because inference is a real, metered, recurring cost that a spreadsheet can find in an afternoon.
Airtable priced AI generously because its founder decided, explicitly, that market share mattered more than margin. That was a rational bet by the people who owned the company. They no longer own the company. Nothing about the arithmetic changed on 4 August — only who is doing it.
The bundled credits in your seat are worth more than the seat. Get that in writing while somebody still has a reason to sign.
Continue Reading
d-Matrix Bought Wallaroo. Get 'Any Hardware' in Writing. Okta Bought Permiso. Your Leverage Expires Oct 31. Copilot's New Billing Turned a $39 Seat Into $750/Month. Anaconda Bought Your AI Red Teamer. Read the License. Tricentis Bought Tabnine's Context Engine, Not Your IDE $60B Bought Cursor. Your Dev Team Is the Product Now. Fixed Budgets, Flexible AI: Atlassian's Answer to Overprovisioned Licenses Gartner's $234B Warning: Will Your SaaS Portfolio Survive 2027?