Of the 44 companies that reported crossing about $100 million ARR in 2026, 33 show how they price. Of those, 18 (55%) charge mainly on consumption: pure usage, credits, a usage hybrid or outcomes. Only 4 rely on seats as their main meter. Free entry is near-universal at the self-serve end, while 11 enterprise-only companies publish no prices at all.
Key takeaways
- 55% of the 2026 $100M ARR class with visible pricing charge mainly on consumption. Only 4 of 33 use seats as the main meter.
- Credit ladders inside monthly tiers are the default for AI apps sold to individuals, from $16 to $250 a month.
- Developer infrastructure publishes per-unit rates and small minimums, like turbopuffer's $16, $256 and $4,096 a month.
- Land and expand is the most common growth engine, led by Linear's reported 177% net revenue retention.
- Treat these figures as reported run-rates. Several mix hardware or one-off revenue, and some are unverified company claims.
In May 2026, a database company called turbopuffer reported that it had reached $100 million in annual recurring revenue.
It had taken 19 months.
And it had raised less than $1 million to get there.
So what does turbopuffer charge? Not a seat. Not a flat plan. It charges per gigabyte stored, per gigabyte written and per petabyte queried, with a monthly minimum of $16, $256 or $4,096 depending on the plan. No free tier.
That's usage-based pricing with a minimum commitment, two ideas straight out of The Subscription Playbook.
turbopuffer wasn't alone. Forty-four companies reported crossing about $100 million ARR in 2026. I went through how every one of them prices, and the pattern is clear: the meter has moved from the seat to the work.
Here's what I found, and what you can borrow for your own business.
Where This Data Comes From (and How Much to Trust It)
The list is based on a chart of 44 companies reported to have crossed about $100 million ARR in 2026. Every company I checked lines up with an ARR Club "crossed $100M ARR" signal in or near the month shown, so it almost certainly reflects ARR Club's 2026 milestone list. I couldn't find the original chart page.
A big caveat before we start. Most milestones are company-reported, often in a LinkedIn post. Arena says outright that its revenue is consumption-based and "not recurring". Pocket's run-rate likely includes hardware sales. ARR Club calls microagi's number unverified.
So read "ARR" on this page as reported run-rate revenue, not audited subscription revenue. The pricing data is more solid: wherever possible it comes from each company's own pricing page, checked in October 2026.
How Do 2026's Fastest-Growing Companies Price?
Here's how the 44 break down by their main pricing model:
| Main pricing model | Count | Companies |
|---|---|---|
| Pure usage / pay-as-you-go | 8 | Tomorrow.io, Kimi, turbopuffer, Arena, Prime Intellect, Deepgram, DeepInfra, Exa |
| Credits bundled into subscription tiers | 5 | Genspark, Emergent, Base44, Venice, Meshy |
| Platform fee + usage (hybrid) | 3 | Hightouch, Hugging Face, Capacity |
| Outcome / per unit of work | 2 | Paraform, Decagon |
| Per seat | 4 | Scribe, Linear, MaintainX, Keepit |
| Flat or modular tiers | 6 | Owner, Luxury Presence, Payhawk, Pliant, Raylo, Monarch |
| Insurance premium | 3 | Lassie, Corgi, Bowtie |
| Hardware + subscription | 2 | Pocket, Tive |
| Not disclosed (enterprise-only) | 11 | Cresta, AHOY, ControlUp, AfterQuery, Legora, Workvivo, Horatio, Kahua, WorldEngine, Skild AI, microagi |
The headline: 18 of the 33 companies whose model is visible (55%) charge mainly on consumption (usage, credits, a usage hybrid or outcomes). Only 4 rely on seats as their main meter.
And even the seat companies are adding meters. Linear charges AI credits for coding sessions on top of seats. Hugging Face combines seats with per-terabyte storage and per-hour compute.
The Full List: All 44 Companies and How They Charge
| Company | Month | Main model | Entry price point (where public) |
|---|---|---|---|
| Genspark | Jan | Credits | Free; Plus from $24.99/month (third-party figures) |
| Emergent | Feb | Credits | Free; Standard $20/month (100 credits) |
| Lassie | Feb | Insurance premium | Quote only |
| Tomorrow.io | Mar | Usage (API) | Free API plan; paid tiers not public |
| Kimi | Mar | Usage (API) | Pay-as-you-go per token |
| Base44 | Mar | Credits | Free; Starter $16/month (annual) |
| Cresta | Apr | Not disclosed | Not public |
| AHOY | Apr | Not disclosed | Not public |
| Hightouch | Apr | Hybrid | Free reverse ETL; paid plans by demo |
| ControlUp | Apr | Not disclosed | Free trial; no prices listed |
| AfterQuery | Apr | Not disclosed | Not public |
| Legora | Apr | Not disclosed | Not public |
| Luxury Presence | May | Flat or modular tiers | On request |
| turbopuffer | May | Usage with minimums | $16/month minimum |
| Tive | May | Hardware + subscription | Quote; free trial trackers |
| Scribe | May | Per seat | Free; Pro Team $13/seat/month (annual) |
| Corgi | May | Insurance premium | Instant online quote |
| Arena | Jun | Usage (evaluations) | Free leaderboard; paid not public |
| Hugging Face | Jun | Hybrid | Free; PRO $9/month |
| Capacity | Jun | Hybrid | Not public (per response, minute or message) |
| Pliant | Jul | Flat or modular tiers | Not public |
| Paraform | Jul | Outcome (per hire) | ~20% to 25% of first-year salary (estimate) |
| Prime Intellect | Jul | Usage (GPU-hours) | Per GPU-hour |
| Payhawk | Jul | Flat or modular tiers | Modules from $299/month |
| Bowtie | Jul | Insurance premium | Instant online quote |
| Owner | Aug | Flat or modular tiers | $249/month + 5% per order, or $499 flat |
| Linear | Aug | Per seat | Free; Basic $10/user/month (annual) |
| Workvivo | Aug | Not disclosed | Not public |
| Aug | Hardware + subscription | $129 device; $200/year plan | |
| Venice | Aug | Credits | Free; Pro $18/month |
| Deepgram | Aug | Usage (per minute) | $200 free credit; from $0.0048/min streaming |
| Horatio | Aug | Not disclosed | Not public |
| Decagon | Aug | Outcome or per conversation | Not public |
| MaintainX | Aug | Per seat | Free; Essential $20/user/month (annual) |
| Meshy | Sep | Credits | Free; Pro $20/month |
| Kahua | Sep | Not disclosed | Not public |
| DeepInfra | Sep | Usage (per token) | From $0.02 per million input tokens |
| Raylo | Sep | Flat or modular tiers | Phones from £6.95/month |
| WorldEngine | Sep | Not disclosed | Not public |
| Skild AI | Sep | Not disclosed | Not public |
| Keepit | Sep | Per seat | Via sales |
| Exa | Oct | Usage (per request) | $10 free credit monthly; $4 per 1,000 searches |
| microagi | Oct | Not disclosed | Not public |
| Monarch | Oct | Flat | $14.99/month or $99.99/year |
Prices checked 27 October 2026; they change often. See the AI Pricing Index for AI companies. Month is when the company reported the milestone in the source chart. MaintainX's 2026 milestone could not be independently verified.
Five Pricing Patterns in the $100M Class
Pattern #1: Credit ladders for AI apps
Most of the AI apps on this list that sell to individuals use a credit ladder: a monthly price that buys a bundle of credits, with bigger bundles at higher tiers. Emergent runs from $20 to $200. Base44 has four paid tiers from $16 to $160. Venice runs from $18 to $200, at 100 credits to $1. Meshy runs from $20 to $100. Genspark goes from $24.99 to $249.99.
Credits let them keep a predictable monthly price while passing model costs through. (More in usage-based pricing for AI.)
Pattern #2: Infrastructure publishes its rates
The developer tools publish per-unit prices and let you start tiny. Deepgram charges per audio minute with $200 of free credit. DeepInfra charges per million tokens, with "priority" at 1.5x and "flex" at 0.8x. Exa charges per 1,000 requests with $10 of free credit every month. turbopuffer adds small monthly minimums so every account covers its costs.
Pattern #3: Outcome pricing is real, but rare
Only two companies price mainly on outcomes. Paraform charges when a hire succeeds. Decagon offers per-resolution pricing, but says most customers pick per-conversation. Buyers still prefer a simple, forecastable meter. (See outcome-based pricing.)
Pattern #4: Free at the bottom, silent at the top
Thirteen companies have a permanent free tier or free product. Others use credit trials (Deepgram's $200) or time trials (Monarch's free first week). At the other end, the 11 enterprise-only companies publish no prices at all. There's very little in between.
Pattern #5: Seats survive by adding meters
The four seat-based companies all keep seats simple and add value elsewhere. MaintainX gives requesters free accounts and charges the people doing the work. Scribe's Team plan costs $13 to $17 per seat against $25 to $35 for an individual. Linear's free plan allows unlimited members.
Which of these models fits your business?
Answer a few questions in the free Pricing Model Picker and see whether flat, seat, usage, credits, outcome or hybrid pricing suits you, with examples.
Land and Expand Is the Growth Engine
The most common growth lesson across these companies is the one I wrote a whole chapter about: land and expand.
- Linear reports 40,000+ paying customers and 177% net revenue retention. On that measure, existing customers grew their spending by a net 77% in a year, before counting a single new customer.
- Legora went from $100 million to $200 million ARR in under six months by expanding inside law firms.
- Horatio says more than 77% of its 125+ clients are growing their teams, and it's fully bootstrapped.
- Payhawk publishes modules "from" $299 to $499 a month, then sells the bundle. Corgi steps its insurance packages up with each funding stage, from seed to Series A to growth.
In Chapter Seven of the book, the first of the five factors for land and expand is an attractive initial offer. Look at the list again: free tiers, $16 minimums, $10 seats. Then the expansion does the heavy lifting. If you want to see what your own retention says about you, run it through the NRR calculator, and read the guide to land and expand.
NRR and GRR CalculatorNet and gross revenue retention.Open the free tool →The Margin Warning Inside the Growth
Fast growth and good margins aren't the same thing.
Base44 is on this list. But according to Ynet's report on Wix's results, Base44's gross margin was "nearly zero" at the start of 2026, with Wix targeting about 60% in the second half after moving to its own model.
That's the AI wrapper problem in one sentence. If your product's main cost is someone else's model, growth can make you poorer. (The guide to pricing an AI wrapper shows how to avoid it.)
Not Everyone Is Selling AI Tokens
Some of the most interesting companies on this list are classic subscription businesses, and they map neatly onto the eight models in Chapter Two of the book.
| Company | What it does | Closest book model(s) | Pricing |
|---|---|---|---|
| Raylo | Rents phones and laptops | Bundled services, with an insurance add-on | Phones from £6.95/month, free upgrade at end of term; 200,000+ active subscriptions |
| Lassie | Prevention-first pet insurance | Insurance + member benefits | Points from app courses worth up to 500 SEK off the premium |
| Bowtie | Online insurer with no agents | Insurance | Instant online quotes; reports retention above 93% |
| Owner | Online ordering for restaurants | SaaS, with the transaction processing pricing model | $249 + 5% per order, or $499 flat; same cost at $5,000/month of sales |
| Monarch | Personal finance app | SaaS (paid only) | $14.99/month or $99.99/year; 1M+ members |
These are the slow builders. Bowtie took about seven years to reach $100 million. Kahua was founded in 2009. Payhawk took five years from $1 million. Meanwhile Pocket reportedly did it in about seven months and Arena in eight months after launching its paid product.
The slow builders are mostly the most verifiable, and their models are the easiest for a normal business to copy.
Run-Rate Is Not Recurring Revenue
One more lesson, and it matters if you ever plan to raise money or sell.
A $100 million "run-rate" built on one-off hardware sales, consumption that may not repeat, or a single great month is not the same as $100 million of contracted subscriptions. Buyers and investors know the difference, and they'll value the business accordingly.
Track your contracted revenue, your net revenue retention and your churn separately. Then see what they imply for your value with the Subscription Business Valuation Estimator.
Free Subscription Profit Review
Australian business owners can book a free 30-minute review with a BPAG Accountants adviser. An honest conversation about:
- whether a subscription model could work for your business
- how recurring revenue could cover your fixed costs each month
- how it could lift what your business is worth
No cost, no obligation. General information only until we understand your situation.
What Can a Small Business Copy From the $100M Class?
- Pick a value metric that grows with your customer's success. Per order (Owner), per resolution (Decagon), per hire (Paraform). For an accounting practice, that might be entities, transactions processed or returns lodged instead of hours.
- Bundle usage into tiers rather than billing raw usage, and add a small minimum so every customer covers their cost.
- Publish your prices and your break-even. Owner's two-plan maths does the selling without a salesperson.
- Make occasional users free and charge the people doing the work (MaintainX, Linear).
- Design the upgrade path on day one (Corgi, Payhawk).
- Give customers a reason to stay at renewal, like Raylo's free upgrade or Lassie's points.
- Watch your margin when you add AI, before you scale it.
What to Do Next
- Choose your own model with the Pricing Model Picker.
- Compare with more AI companies in AI pricing examples and the live AI Pricing Index.
- Check your numbers against the 2026 subscription benchmarks.
Frequently asked questions
Where does this list of companies come from?
It's based on a chart of 44 companies reported to have crossed about $100 million ARR in 2026. Each company I checked matches an ARR Club 'crossed $100M ARR' signal in or near the month shown. I couldn't find the original chart page, so treat the list as a strong sample of the 2026 class, not an official ranking.
Are these ARR figures audited?
No. Most milestones are figures the companies reported themselves, often in a LinkedIn post or funding announcement, then summarized by ARR Club or the press. Arena says its revenue is consumption-based and 'not recurring', and ARR Club itself calls microagi's figure unverified. Read 'ARR' here as reported run-rate revenue.
Which company on the list grew fastest?
Among the reported figures, Pocket went from $1 million to $100 million run-rate in about seven months, and microagi claimed $100 million six months after founding, though that claim is unverified. Arena reached $100 million eight months after launching its paid evaluations service. Much of Pocket's run-rate is likely hardware sales rather than subscriptions.
Do any of these companies price like a normal small business could?
Several do. Owner charges restaurants $249 a month plus 5% per order, or $499 flat. Monarch's core plan is a flat $14.99 a month or $99.99 a year. Raylo rents phones from £6.95 a month. Those structures, a flat plan, a choice between flat and per-transaction, or rent-instead-of-buy, work for local businesses too.
How often is this page updated?
This edition reflects research checked on 27 October 2026. Prices and milestones change quickly, and new companies cross $100 million every month. For the latest prices of the AI companies on this list, check the AI Pricing Index and each company's pricing teardown, which carry their own check dates.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- ARR Club, turbopuffer ARR hit $100M in 19 months with less than $1M raised (May 2026)
- turbopuffer pricing (checked 27 Oct 2026)
- ARR Club, Linear hit 40k paying customers with a 177% net revenue retention rate (Aug 2026)
- ARR Club, Pocket ARR hit $100M from $1M in just 7 months (Aug 2026)
- TechCrunch, Arena is now a $100M business (29 Jun 2026)
- Ynet, Wix and Base44 results (Aug 2026)
- Owner pricing (checked 27 Oct 2026)
- ARR Club, Legora ARR surpasses $200M (2026)
- ARR Club, Horatio ARR surpasses $100M fully bootstrapped (Aug 2026)
- ARR Club monthly rankings (Aug 2026)
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.