What is your company's data worth to AI labs?
The value of company data to AI labs depends mostly on four things: how many people produced it, how many years it covers, which systems it comes from, and whether the license is exclusive. This calculator turns those into an illustrative first-year range. It is not an offer.
Illustrative first-year range
$170K to $395K
An estimate from seat-years of history and the systems you selected, kept inside the ranges buyers publish. It is not an offer. A real number follows a rights review. How this is calculated.
Five minutes. No call required.
Method
How the estimate is calculated.
No black box. The formula is simple on purpose, and it is only a starting point.
The estimate multiplies seat-years of history (employees times years of records, with years capped at ten) by a rate for each system you select. A refreshable dataset adds 25 percent. An exclusive license adds 50 percent. The range shown is 60 to 140 percent of that figure, held between $10K and $5M, the span that buyers of company data publicly advertise.
The rates are Outrich Data's own working assumptions, set so that mid-size companies land inside those published ranges. They are not market prices. A real offer depends on the data title search, the quality of the records and what buyers need at the time.
Rate per employee, per year of records
| System | Rate |
|---|---|
| Email and chat | $60 |
| Documents | $40 |
| Support tickets | $50 |
| Call recordings | $80 |
| Source code | $100 |
| CRM and ERP records | $40 |
| Video and images | $60 |
| Workflow logs | $50 |
Ranges buyers publish
| Buyer | Published range |
|---|---|
| Scale AI | $10K to $1M+ per partnership (illustrative) |
| micro1 | $100K to $3M+ |
| Mode | $100K to $5M |
| Polyshares | $100K to $2M and above |
| DataPaid | $100K to $2M per company |
Sources
- Scale AI, as published on its own site2026-11-16
- micro1, as published on its own site2026-11-16
- Mode, as published on its own site2026-11-16
- Polyshares, as published on its own site2026-11-16
- DataPaid, as published on its own site2026-11-16
For how buyers think about price, read AI training data pricing and how to value a dataset.
Find out if your data qualifies.
Eleven questions, about five minutes, no call required. You get an answer on screen and a written follow-up.