Firebase Firestore
Google · databases · reviewed by Stan K. · Source: data snapshot · data as of 2026-11-12
Per-document pricing where the free tier caps by the DAY (50K reads/DAY) and Blaze has no hard spend cap — requests “will succeed even if you have exceeded your monthly budget”, Google’s own docs.
#8 of 9 in Databases
ToolScore breakdown — why 54/100
10 criteria × 10 points. Low pillars are visible on purpose — that's the point of a score you can audit. Methodology →
What it does well
- ▸$0 if traffic is even — ≈33K reads/day sits under the 50K reads/DAY free cap
- ▸The free caps are daily, not monthly: a bursty day doesn’t wait for a monthly reset
- ▸Per-document unit math is public: $0.03 reads / $0.09 writes per 100K (us-central1), nam5 at exactly 2×
- ▸The spend-cap behavior is documented in Google’s own words — the fine print at least speaks plainly
Where it costs you
- ▸Blaze has NO hard spend cap: requests “will succeed even if you have exceeded your monthly budget” — Google’s own docs
- ▸Security Rules get()/exists() are billed reads — your auth layer is a read amplifier
- ▸nam5 multi-region = exactly 2× the regional rates ($0.06 read / $0.18 write / $0.02 delete per 100K)
- ▸Free daily caps apply to the default database only — named databases get none
- ▸The exit: Firestore → SQL is a 2–6 week rewrite ($5–25K+)
Hidden traps — the stuff the pricing page doesn't say
Google’s own docs: requests “will succeed even if you have exceeded your monthly budget”. The bill-shock ladder from our hub: 5K DAU loading a 500-doc catalog with no pagination turns a $30/mo app into an $800/mo app; a runaway function loop goes $10 → $4,976; the community ceiling is a $50 bill that reached $70,000. Nothing stops it for you — paginate every query, cache aggressively, alert on daily read counts, and write your own kill switch.
Security Rules get()/exists() are billed reads: every rule that checks another document adds a read to the bill. The layer that protects the data is priced like the layer that reads it.
Firestore → SQL is a 2–6 week rewrite ($5–25K+), against Supabase’s pg_dump exit in 1–5 days. If you can’t afford to leave, you don’t own the architecture.
The nam5 multi-region configuration is exactly 2× the regional rates: $0.06 read / $0.18 write / $0.02 delete per 100K. Multi-region is a pricing decision as much as an availability one.
The free daily caps (50K reads/DAY + 20K writes/DAY) apply to the default database only — named databases get none. Split your data across databases and the $0 tier silently evaporates.
Know the limits before you pay: read the traps above, then decide with eyes open.
Visit official siteAffiliate link where available — never affects the score.
- official Firestore pricing — official, checked 2026-11-12 verified
- official Firebase pricing (Blaze, spend-cap behavior) — official verified
Data as of 2026-11-12. Prices and limits change monthly — report anything stale. How we verify: methodology.
Closest alternatives
No free tier and no $0 ramp — day one is PS-5 at $5/mo ($15 HA 3-node) — and the engine is Postgres
Hourly cluster tiers with a rare honest cap — Flex ($8–30/mo, 5GB) publishes its monthly ceiling — b
Compute billed in CU-hours with scale-to-zero after 5 min — $0 for idle, ≈$19.35/mo always-on (0.25
Firebase Firestore review
In-depth analysis · last updated 2026-11-12
Cheap if flat, unbounded if spiky
Firestore bills per-document operations, and at an even pace it is the cheapest thing in this comparison: our hero workload (1M reads + 100K writes) breaks down to ≈33K reads/day — under the 50K reads/DAY free cap — so the month is $0. The catch is in the denominator: the free caps are DAILY, spike days over them are billed, and every read above the cap carries the us-central1 rate. The spiky version of the same month costs $0.39 — which reproduces from the page as 10 × $0.03 reads + 1 × $0.09 writes per 100K.
The fine print that runs the bill
Three lines of fine print do more damage than the unit price. First, Blaze has NO hard spend cap — Google’s own docs say requests “will succeed even if you have exceeded your monthly budget”. Our hub’s bill-shock ladder: 5K DAU loading a 500-doc catalog with no pagination turns $30/mo into $800/mo; a runaway function loop goes $10 → $4,976; the community ceiling is a $50 bill that reached $70,000. Second, Security Rules get()/exists() are billed reads — the auth layer is a read amplifier. Third, nam5 multi-region is exactly 2× the regional rates ($0.06 read / $0.18 write / $0.02 delete per 100K), and the free daily caps apply to the default database only — named databases get none.
The exit is the real price
Against Supabase’s pg_dump exit (1–5 days, $500–3K), Firestore → SQL is a 2–6 week rewrite at $5–25K+. The per-document model that makes the entry free makes the leaving expensive — if you can’t afford to leave, you don’t own the architecture.
How the pillars land
Price/Value 6.5: $0 if even, $0.39 if spiky — the bill tracks your worst day, not your average. Real limits 6: 50K reads/DAY + 20K writes/DAY + 1GiB + 10GiB egress/mo, with the named-database carve-out. Latency & cold starts sits at the neutral 5 — no latency or cold-start facts in this snapshot. Data handling sits at the neutral 5 — no data-handling facts in this snapshot. Portability & export 3: the 2–6 week, $5–25K+ rewrite to SQL — the worst exit fact in this table. Feature set 6.5: per-document model, Security Rules, named databases, multi-region. Track record sits at the neutral 5 — no track-record facts in this snapshot. Ease of use 5, neutral — and the rules-billing trap actively works against it. Pricing transparency 6: unit rates are public, but the no-cap behavior and the named-database carve-out are the parts that matter. Support & docs 6.5: Google’s own docs state the no-cap behavior plainly enough that this page can quote them verbatim.
Verdict — NO SPEND CAP, 54/100
Firestore is the right database for flat, read-light, document-shaped products — and the wrong one to meet your first viral day on. Use it inside guardrails you build yourself: pagination on every query, aggressive caching, alerts on daily read counts, and a kill switch. Blaze will not stop for you; the docs are honest about that, and so is the score.