COMPARISON · HOSTED MODERATION APIs

Flat cost,
not per character.

Google Cloud's Text Moderation bills per unit of characters — a cost that grows with every message. Filter8's marginal cost is compute, so it stays flat as you scale, runs sub-millisecond, and keeps nothing after the response. (This is the paid Cloud NL classifier — not the free Perspective API, which is being retired.)

Try the playground → See flat pricing
100 free calls · no per-character metering · run the math on your volume.
Flat, not per-character
A per-character-block price is linear in volume — long messages cost more, every message forever. Filter8 sustains huge throughput per core, so cost barely moves as you grow.
Sub-millisecond, in-process
A hosted classifier is a network round-trip on every message. Filter8 decides in microseconds, fast enough for a synchronous send-time path.
Deterministic & PII-aware
A replayable verdict plus built-in phone/email/URL classifiers — including spelled-out — and nothing retained after the response.

Google Cloud Text Moderation vs Filter8

CAPABILITY Google Cloud NL Filter8
Cost modelPer character block — linear in volumeFlat — marginal cost ≈ compute
LatencyHosted round-tripSub-millisecond, in-process
ApproachHosted ML classifierDeterministic engine
OutputCategory confidencesVerdict + clean_text
Explainable / replayableScores✓ reproducible explain trace
Data residencyText sent to GoogleNothing retained after the response
Evasion (leet · homoglyph · spacing · zero-width)Model-dependent26 classes, deterministic
Contact / PII (incl. spelled-out)Separate APIBuilt-in classifiers
Rating-aware (E–M · minors)Per-rating, per-category ceilings
Multilingual ML breadth— English-first, deterministic
Google Cloud Natural Language is a Google Cloud product; its Text Moderation is distinct from the (retiring) free Perspective API. Comparison as of July 2026 based on public information and Filter8's positioning; verify current Google Cloud rates and features before relying on specifics.
THE MATH THAT MATTERS

Per-character is a tax that scales with your success.

A per-character-block price looks trivial on a demo and compounds at scale — it's multiplied by every message, and longer messages cost more. A single core sustains tens of thousands of messages per second, so volume that runs into five or six figures a month on a per-call API is a rounding error in compute for a deterministic core.

Note the migration context too: the free option many teams used — Perspective — is being retired, which is what pushes teams toward paid hosted classifiers like this one. A deterministic, flat-cost core is both cheaper and more durable than swapping one per-call vendor for another.

WHERE GOOGLE CLOUD NL WINS

Honest about the trade

Google Cloud Natural Language brings broad multilingual ML and deep GCP ecosystem integration, managed and maintained by Google. If you're all-in on Google Cloud and need multilingual nuance or the wider NL toolkit (entities, syntax, sentiment), it's a natural fit. Filter8 is deliberately narrow: the fastest, cheapest, most private path for high-volume text moderation, and it escalates the ambiguous tail to a model when you want that nuance.

Google Cloud NL comparison questions

Isn't this the same as Perspective?
No. Perspective is Jigsaw's free toxicity API, and it's being retired. Google Cloud Natural Language Text Moderation is the paid Cloud classifier. Both are hosted ML and a network hop; Filter8 is deterministic, flat-cost, and keeps nothing after the response.
How much cheaper is Filter8?
It scales with volume. Google bills per character block; our marginal cost is compute, so the gap widens the more (and the longer) the messages you filter.
Is deterministic accurate enough?
For known patterns and evasion on the hot path, yes and far faster. Route the ambiguous or multilingual tail to a model asynchronously.
Where does my text go?
Managed keeps nothing after the response and logs offsets, not substrings — nothing is handed to a third-party model.
COMPARE MORE

Filter every character, flat.

No per-character meter to ration against. 100 free calls, no card.

Get an API key Open the playground