Firestore's free quota is daily, not monthly — here is why Blaze bills spike
Moving from Spark to Blaze does not turn off the free quota; it adds metered billing on top of the same daily numbers. A page that re-queries a whole collection on every load, or a feed that re-renders on every keystroke, can burn through a day's free reads before anyone notices a bill is accruing.
Last checked November 18, 2026.
Firestore's free quota is daily
That table resets daily, not monthly, and the Blaze plan (the only plan that unlocks most of Firebase's other services, including outbound network requests from Cloud Functions) keeps the same free daily quota underneath — Blaze does not remove it, it adds metered billing for whatever crosses it. The exact per-100,000-operation rate past the quota is on Google Cloud's own Firestore pricing pages, not reproduced here since this page was not fact-checked against that specific rate table.
| Operation | Free quota per day |
|---|---|
| Document reads | 50,000 |
| Document writes | 20,000 |
| Document deletes | 20,000 |
Sources: Firebase: Firestore pricing
The common way a bill spikes
- A listing or dashboard page that fetches a whole collection on every load (instead of paginating or caching) multiplies its read count by every page view, every user, every day.
- A live search box that queries on every keystroke turns one search into many reads, each one counted separately.
- A recursive delete of a large collection (deleting a parent document does not delete its subcollections automatically) can undercount what is actually being billed if the app assumes one delete removed everything underneath it.
- A missing composite index forces a query to read more documents than it returns, so the bill reflects documents scanned, not just documents matched.
Sources: Firebase: Firestore pricing
SQL has no per-document meter
Postgres has no per-read or per-write operation meter at all — a query that scans a million rows and returns ten costs compute time, not a line item per row touched. Every Scribase project is flat-priced ($5 a month on Hobby, $25 on Pro) with no equivalent to Firestore's per-operation billing; a badly written query is a performance problem to fix with an index, not a line item that keeps growing on its own.
The Firebase importer moves Firestore collections into Postgres tables with a jsonb data column, so an app rewriting its hottest, most-read collection away from per-document billing keeps querying with SQL from day one instead of re-deriving the schema from scratch.
Sources: Scribase pricing
Questions people ask
Does the free Firestore quota carry over if I do not use it one day?
No. Firebase's own documentation states the quota as a daily allowance; an unused quota on one day does not add to the next day's allowance.
Is the free quota the same on Spark and Blaze?
Yes, per Firebase's pricing docs — the daily free quota (50,000 reads, 20,000 writes, 20,000 deletes) is the same number on both plans. Blaze adds metered billing past it; it does not raise or remove the free allowance.
Can I set a hard spending cap so Firestore billing cannot exceed a number I choose?
Firebase's own pricing documentation does not describe a built-in per-service hard spending cap on this page; budget alerts through Google Cloud Billing notify you, but do not themselves stop usage once a threshold is crossed.
Does deleting a parent document delete its subcollections too?
No — this is a known Firestore behavior, not specific to pricing: deleting a document does not automatically delete documents in its subcollections, which can leave orphaned data still counting against reads and storage until it is deleted explicitly.