Things researchers ask first.
Data privacy, citation, how tokens are used and restored, dataset size, and academic pricing. If your question isn't here, ask. A person answers.
Will my data be used to train a model?
No. Your data is processed only to produce your report, kept in your account alone behind an ownership check on every file and figure, deleted on a fixed retention window, and never used to train anything. There is no research exception and no academic loophole.
Can I cite this in a paper?
Yes. Completed analyses include Methods text describing the tools and analysis settings used, with references for citation. Review it alongside your results and adapt it to your manuscript. See an example Methods paragraph →
Does submitting a run spend my tokens?
Submitting reserves tokens, reducing the amount available for other runs. A completed run uses those tokens. Some other outcomes also use them, as explained below. If processing never starts, the reserved tokens are restored.
Do tokens expire?
No. Tokens stay on your account until used. There is no subscription and no monthly minimum.
Are there feature differences between paying users?
No. We do not sell premium or lite versions. Analysis options depend on the assay, configuration and data. Dataset size sets the token count.
What if the run fails?
Failures on our side, including crashes, memory errors and processing timeouts, restore the reserved tokens automatically. If processing has started, validation failures and corrupt input use the reserved tokens. Cancelling after processing starts or letting a required review expire also uses those tokens.
How big can my dataset be?
Up to one million cells per self-service sequencing run. Larger cohorts, non-standard designs, and organisms beyond human and mouse route to a Custom Analysis brief, scoped and quoted before any work starts, billed at $300/hour academic / $500/hour non-academic.
What if an analysis can't run on my dataset?
Some analyses depend on the data and available controls: trajectory inference needs enough cells, differential comparisons need a suitable study design, and spectral unmixing needs reference controls. The report records which analyses ran and which were unavailable. If a required analysis cannot complete, the run reports a failure.
What counts as my "input size"?
Cells for sc/snRNA; events for cytometry; samples for bulk. It's locked once the platform's first pass parses your input: no surprise re-pricing mid-run, and the number you see on the upload page is the number you pay.
How does academic pricing get verified?
Automatically, the moment you verify an institutional email. It is granted on verification rather than self-declared, so nobody is taken at their word. If your institution isn't auto-recognised, you request academic pricing in the app and we review it.