ds@mcpanalytics.aiA data scientist you can email your data and question to, and get a reliable response. Not AI slop.A data scientist you can email. Not AI slop.
Credits are the one unit for everything — running an analysis, and the review that writes it up. What each costs moves with the model that reads your results and with what the work takes, so we do not quote a fixed rate. You are never charged for a run that fails, and we check you can afford one before it starts. Refreshes cost half. Failed builds cost nothing.
| item | credits |
|---|---|
| Welcome grant | +1,000 |
| Churn drivers — run | as charged |
| A build that failed on our side | −0 |
| Churn drivers — re-run | as charged |
A credit is our unit of pricing. When you run an analysis, it costs credits. The cost is calculated automatically from how much AI processing and compute the module actually uses in production. You always see the credit cost before running an analysis.
Prices are flat and shown before you run — no formulas to decode. Here is a month, itemized the way your ledger actually reads:
| item | credits |
|---|---|
| Welcome grant (signup) | +1,000 |
| Sales trend — run and write-up | as charged |
| Churn drivers — run, read closely | as charged |
| A build that failed on our side | −0 |
| A second question, answered | free |
| Retention study — re-run next month | as charged, no build fee |
The ledger is append-only — every line explains itself, and you approve the spend before anything runs.
Simpler analyses with fewer insight cards cost less. Complex multi-section reports with longer compute cost more. Here's the general range:
The exact credit cost for every module is shown before you run it — in the tool discovery results and on the confirmation screen. No surprises.
Run out before the month is up and you can buy a boost: one seat's allowance again, at the seat's own price, for whichever designation you are on. It is the same credits doing the same work, bought mid-cycle rather than waiting for the refill.
| Boost | Price | Credits | Window |
|---|---|---|---|
| Academic boost | $20 | 2,000 | 30 days |
| Business boost | $20 | 2,000 | 30 days |
Every plan keeps your reports archived permanently.
| Plan | Price | Credits/mo | Credits per seat | What stands out |
|---|---|---|---|---|
| Academic | $20/mo | 2,000 | 1 seat | One seat: one person, with a month of usage on it |
| Business | $400/mo | 40,000 | 20 seats | A standing analyst for a team |
Every module's credit cost is calculated from its real production data and re-pegged periodically to the linear regression baseline. Today, the range looks like this:
You always see the exact cost before you run anything. And creation is only charged when the new module reaches the complete state — failed builds are absorbed on our side.
Module credit costs are recalculated periodically based on actual production data. As we optimize a module (fewer AI calls, faster execution), its cost may decrease. If a module gets more complex (additional insight sections, deeper analysis), its cost may increase. The formula is automatic — costs are always pegged to the linear regression baseline.
Every report has a "Report an Issue" button. If a report doesn't deliver — a chart didn't render, the analysis wasn't useful, or something went wrong — click it and you get a doOver: a free re-run of the same analysis, consumed before any credits are touched. Try different parameters, adjust your data, run it again at no cost. If the re-run still doesn't meet your expectations, you can request a refund — reviewed case by case. Failed builds are never billed.
Demo accounts get every model and every platform tool for 14 days, no card, subject to fair use. Fair use is a published allowance, not a mystery: each demo account receives a month's worth of usage for the demo period. The demo does not have its own price list: everything costs on the demo exactly what it costs on a paid plan — the demo has no separate price list. That is room to ask real questions of real data for a fortnight, or 450 runs of analyses you already own. Reaching the allowance never ends your demo: your remaining days, your reports, and your data stay available, and you can ask us to extend it.
Cost-conscious or high-volume? Register your own Anthropic API key and the AI narration on your analyses runs on your token spend, our software. Available today via the API (POST /api/v1/user/llm-keys — validate, store, delete); account-page management is coming. Your key is used only for your own work.
Without a plan, reports are kept for 1 year. But once you cite a report — generate a citation for it, share it via a link, pin it, export it to PDF with methodology, or reference it in another analysis — it switches to permanent retention regardless of your tier. Cited reports represent commitments (to a paper, a board deck, a client deliverable) and we will not quietly delete them. This is core to the "reproducible analysis archive" product promise.
On the paid plans, every report is retained forever whether it's been cited or not.
Where to start
One email with your data and your question. We take it from there.
Cymple
Data Scientist
Send me your data and question, I’ll send you the analytics.
Try it on your own data
14 days free.
The whole platform, every depth. No card, no call, nothing to install.
14 days free. →Tell us what went wrong, in your own words. We capture the page you're on automatically, so no need to describe where you are.