Fifty dollars a month.
For the practice, not per case and not per administration — you already pay those to the people who own the norms. Colleagues are $25 each.
plus $25 a month for each additional person who signs in. Or $500 and $250 a year — ten months’ money for twelve.
- Unlimited cases, scores and instruments
- Report material for every case, to print or paste
- Cohort comparison across your own series
- Report and vendor-export import
- Full export, whenever you want it, in a documented format
- 14 days to try it, without a card
What a practice actually pays
| People signing in | Monthly | Annually |
|---|---|---|
| Just you | $50 | $500 |
| 2 clinicians | $75 | $750 |
| 3 clinicians | $100 | $1000 |
| 5 clinicians | $150 | $1500 |
Counted rather than chosen: it follows who can actually sign in, so there is no seat to remember to buy and none to remember to give back. A colleague added mid-month appears on the next invoice, not as a charge that afternoon.
What it is worth comparing against
Not other software. The hours after a battery — typing a score table out of several portals, then writing the paragraph about it. If Casebook saves an hour a case and you see two cases a week, it has paid for itself several times over before you have thought about it again.
Why there is a trial and not a free tier
14 days, no card, because you cannot tell whether this is worth paying for without running a few real batteries through it. A permanent free tier is different: in a field this small it produces goodwill and no signal, and a practice that will not pay fifty dollars a month will not tell us anything true about whether this is worth building.
What happens if you stop
You export everything first — a zip of CSVs with a codebook describing every column, which is the same export available to you on any ordinary day. Then closing the practice deletes it all; encrypted backups age out within thirty days. Nothing is held hostage to a renewal.
What is not in the price yet
Scanned reports. If a report is an image with no text layer, Casebook refuses it rather than half-reading it, and says so. Optical character recognition is the obvious next thing and it is not built.
Still deciding?
Read what it stores and what it refuses to store. It is the part most worth arguing with.