Misata Logo
Misata Studio
Beta: temporary pricing proposal

Free by invitation.
Pay for the work, not the rows.

Studio is invite-only for now, and invitations go out from the waitlist. Every account gets 20 credits a month without a card, which is 2 complete builds: describe it, review the design, get verified data. Studio is in beta, and this is a temporary proposal: prices and the free allowance may change while we learn what the work really costs. A failed run is refunded automatically.

Join the waitlist Or download free datasets
Freeevery month, in beta

20 credits, enough to build 2 datasets of up to 50k rows each. Unused free credits do not carry over.

  • 20 credits a month: 2 complete builds, from a sentence to verified data
  • Every guarantee, certificate, export and the review-before-generate step. Nothing is held back
  • Correct data by prompt, ask it questions, share a certificate
  • One design or build at a time, and a few an hour, so the free tier stays free for everyone. Buying credits lifts these limits
  • No card, now or later

A failed run costs nothing. Design that cannot be made, data that does not verify, a correction that cannot be applied: refunded on the action itself, without asking.

Bring your own key and pay only for rows. Then we are not running the model, so design is free and a build is one credit per 10,000 rows.

What things cost

The price follows the model work, because that is what costs real money and it costs about the same for 500 rows as for 50,000. A complete build is 10 credits.

ActionCredits
Design the tables2
Make the data8
Each further 10,000 rows1
Correct a dataset by prompt3
Your own model key1 per 10,000 rows

Credit packs

One-off, no subscription, and they do not expire. Buy one when a project needs more than the free credits.

Starter
$9
50 credits
5 complete builds, about $1.80 each

A side project, or trying Studio on real work.

Studio
$29
200 credits
20 complete builds, about $1.45 each

Everyday work on staging databases and demos.

Team
$99
900 credits
90 complete builds, about $1.10 each

Large schemas, daily seeding and a team sharing one balance.

Packs are bought from inside the app, where your balance lives. These are temporary beta prices and may change.

Will the free credits cover you?

Set your real workload. Designing the tables costs 2, making the data 8 (up to 50,000 rows), and every account gets 20 credits a month for free.

per dataset 10 creditsper month 120that is 360k rows

You would need 100 credits beyond the free 20 each month. That is one Studio pack, $29 (about $1.45 a dataset), and packs do not expire.

Credits are only spent on work that arrives. If a run fails, the credits go back automatically, and you can see every movement in your account.

How others charge

Read from their own pricing pages on 21 September 2026. Plans change, so check the source.

  • Tonic Fabricate: Free with $5 of monthly usage; Plus $29 a month with $25 of usage, then pay as you go. Billed by AI tokens.
  • Seedfast: Free with $5 of monthly credit (about 8 runs on a 30-table schema); $16 and $69 a month with monthly credit that does not roll over.
  • Mockaroo: Free up to 1,000 rows per file; $60 a year up to 100,000 rows; no model in the loop.

The tools that put a model in front of your data (Fabricate, Seedfast) charge by the model work, not the row. That is the model here too, with the price of a complete build shown up front.

Questions

What is a credit?
A unit of work. Designing the tables costs 2, making the data costs 8 (the first 50,000 rows included), and every further 10,000 rows cost 1. So a complete build is 10 credits. Every account gets 20 a month, which is 2 builds. Unused free credits do not carry over to the next month; credits you buy never expire, and free ones are always used first.
Why is a small dataset not much cheaper than a bigger one?
Because the expensive part is the model work: understanding the request, designing the tables, deciding what every column means, and checking the result. That costs about the same for 500 rows or 50,000. The rows themselves are cheap, so past the first 50,000 they are one credit per 10,000.
Why is the free tier smaller than before?
It used to be a million free rows a month, which measured the wrong thing. It is now two complete builds a month, enough to find out whether Studio does what you need, in a product that now runs a model on your behalf. This allowance is part of the temporary proposal and may change.
What does beta mean for pricing?
Studio is in beta: things change and some edges are rough. These prices are a temporary proposal, not a final price list: the free allowance, what a credit buys and the pack sizes may change while we learn what the work really costs, and we will say so on this page when they do. If something fails you are refunded automatically, and the limitations are listed rather than hidden. Read the limitations page.
What happens if a run fails, or the design could not be made?
The credits come back automatically, on the same action, without you asking. If the design model does not answer in time, the design step is refunded. If a correction cannot be applied, or does not pass verification, it is refunded. Every movement is listed in your account, so the balance can always be explained.
Can I use my own AI key?
Yes. Add a provider key in Settings and design is free and a build costs only its rows, one credit per 10,000 rows. The key is sealed to your account and never readable by us. We run the model otherwise, and that is what the credits pay for.
Do credits expire?
Bought credits do not expire. The free monthly credits are added each month rather than resetting your balance, so an unused month is not lost and buying credits never wipes them.
Why credits instead of a monthly subscription?
A subscription is a promise that the service is available all month. Studio is young enough that promising units of work delivered is more honest than promising uptime. When that changes, a subscription can too.
Is anything held back from the free tier?
No feature is. Every guarantee, certificate and export works on the free tier. Credits meter the model work and the volume, not capability, because a guarantee you can only afford at the top tier is not much of a guarantee.
Is the open-source library free?
Yes, always. The MIT-licensed engine runs from the terminal with no account and no limit. Credits pay for us running Studio and its models for you, not for permission to use the engine.

Need something the packs do not cover?

A dataset built to your specification, a volume beyond the largest pack, or a guarantee the language cannot express yet. Tell me what you are trying to make true and I will tell you honestly whether Studio can do it, including when the answer is no.

Ask for it Custom dataset service

On being early

Studio is built by one person, and it is early enough that the people who pay for it now are doing something more than buying credits. They are funding the hours that go into the engine, and they are the clearest signal I get about which guarantee to build next.

Buy credits because the tool does something you need, not to be kind. If it does not do that yet, tell me what is missing instead. The free credits are there so you can find out before spending anything.

The library is free, and always will be

The MIT-licensed engine runs from the terminal with no account and no limit. Credits pay for us running Studio and its models for you, not for permission to use it.

pip install misata
misata generate --config schema.yaml --output-dir ./data
Read about the engine