Updated 30 August 2026 · one-time payment, no subscription
Buy a qpAdm analysis: a hand-checked model from your raw DNA
qpAdm itself is free software. What you buy here is the analysis: your genotypes merged into AADR v66, a model composed, run and checked by a person, and a report with the p-value, standard errors and z-scores that say how much to trust it. From €29.99, one-time.
Formal model report
Model test
Source confidence
Model record
What you are buying
You cannot buy qpAdm: it is the qpadm() function of ADMIXTOOLS 2, published by David Reich's laboratory and free to anyone who can run R. Every service that sells "qpAdm" sells one of two things: a do-it-yourself environment where you build the models, or an analysis where someone builds and checks them for you. Ancestrify sells the second, and adds the first afterwards as an optional unlock.
- Price
- From €29.99 to €59.99, paid once. The tier sets how long the analyst searches, never the quality bar.
- Input
- A raw-data export from 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA or LivingDNA, or a whole-genome VCF (+€10)
- Reference panel
- Allen Ancient DNA Resource (AADR) v66, merged with your own genotypes
- The model
- Composed, run and reviewed by a person; only a model that passes the publication bar (p-value, every z-score, every standard error) is published
- Statistics you receive
- The model p-value; each source's weight, standard error, z-score and 95% confidence interval; the full right set with sample counts; the complete run record as plain text
- Afterwards
- The Model Lab (€10.00, one-time): up to 100 of your own qpAdm runs a day on your own merged dataset, plus the EIGENSTRAT bundle to download
Four depths, one report
Every tier delivers the same report and every published model passes the same check. What changes is how many models a human analyst builds, compares and stress-tests before publishing your strongest one. Medium is pre-selected at checkout.
- The focused searchBase€29.99
Your strongest straightforward model — found, checked, published.
≈ a few hours of analyst time · 15–25 hand-built models
- The search past good enoughMedium€39.99
We keep searching past the first model that works — until a clearly better one stops turning up.
≈ about one working day of analyst time · 50–70 hand-built models
- The full sweepDeep€49.99
Every plausible version of your ancestry tried — the model that survives being pushed.
≈ several working days of analyst time · 80–120 hand-built models
- The exhaustive searchPerfect€59.99
The search ends only when nothing beats it — the strongest model your DNA can give.
≈ a week or more of analyst time · 120–200 hand-built models
How ordering works
- Check your file for free. The file check reads your export in the browser and shows the coverage verdict before you create an account.
- Create an account and upload the file. Choose the tier and any add-on, consent to the processing, and pay once by card.
- We merge your genotypes with the AADR v66 panel on our own servers.
- An analyst composes, runs and checks the models by hand, exhausting the candidate sources and their alternative labels.
- Your report publishes in your private dashboard, and you are emailed. The Model Lab is then one unlock away.
Accepted files and refunds
Standard autosomal exports are accepted as .txt, .csv, .zip or .gz up to 50 MB; a whole-genome .vcf or .vcf.gz up to 1 GB is converted to the panel's markers at upload and not stored. An unreadable or truncated file is refused before payment. The refund policy sets out the cases where an order is refunded, including when no report can be delivered.
Other ways to get qpAdm
If you would rather build the models yourself, the free AdmixTools 2 Lab runs real f-statistics, qpWave and qpAdm over the public reference panel in the browser, with no account. After a published report, the Model Lab does the same with your own sample as the target. Our guide to every qpAdm service lists the do-it-yourself and analyst-built options side by side, and the comparison pages put the facts in one table.
Questions
Can I buy qpAdm itself?
No, and nobody can: qpAdm is free, open-source software, the qpadm() function of ADMIXTOOLS 2 from David Reich's laboratory. What you buy here is the analysis around it: your raw genotypes merged into the Allen Ancient DNA Resource v66, a model composed, run and checked by a person, and a report that publishes the p-value, the weight, standard error and z-score of every source and the full right set.
How much does a qpAdm analysis cost?
From €29.99 to €59.99, paid once. The four tiers deliver the same report and pass the same quality check; a deeper tier buys a longer human search past the first model that passes. Optional add-ons are also one-time: the Model Lab (€10.00) to run your own models afterwards, and the whole-genome VCF upload (€10). There is no subscription and no credit system.
Which tier should I choose?
Medium is the tier we recommend and pre-select: We keep searching past the first model that works — until a clearly better one stops turning up. Base is right if you want the strongest straightforward model at the lowest price; Deep and Perfect are for someone who wants every plausible version of their ancestry tried before the model on the table is called final. The quality bar never moves between tiers.
Which files can I upload?
A standard raw-data export from 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA or LivingDNA (.txt, .csv, .zip or .gz, up to 50 MB), or a whole-genome sequencing VCF (.vcf or .vcf.gz, up to 1 GB) with the €10 add-on. Check a file for free first: the file check reads it in your browser and tells you, per product, whether it can be used.
Do I need an account, and how do I pay?
Yes, a free account: the report is private to you, and the order, the analyst's messages and the later Model Lab all live in it. Payment is by card at checkout, in euros, once. You upload the file, choose the tier and any add-on, consent to the processing, and pay; nothing is charged before the file has passed the pre-payment checks.
What if no model passes the quality check?
It happens, and it is the method working: qpAdm can reject every candidate. The analyst reports exactly that, with the closest models and why they failed, rather than publishing a weak fit dressed as a result. The refund policy covers the case where we cannot deliver a report; the Model Lab remains available on any published report for exploring alternatives yourself.
Is this real qpAdm or an approximation?
Real qpAdm: the qpadm() function from ADMIXTOOLS 2, run in R on our own infrastructure over your genotypes merged with AADR v66. Nothing is simulated from coordinates, and no automated rotation picks the model; the report's Reading view shows the complete run record so a reader who knows the method can reproduce it.
A model that can say no, checked by a person.
Upload the raw file you already have, choose the depth of the search, and receive a formal qpAdm model with the statistics that let you judge it.




