For when checking should happen automatically — at signup, at checkout, or as a nightly sweep over your CRM — rather than someone uploading a spreadsheet by hand. No WhatsApp account is involved, so nothing in your stack carries restriction risk.
Worth being direct before you plan an integration around it.
Turbo Filter drives a logged-in WhatsApp Web session. There is no endpoint to call, and exposing one would not help — the bottleneck is an account that can be restricted at any point. A pipeline built on that fails unpredictably in production.
Validation runs server-side with no WhatsApp login: no account to warm up or replace, no Windows machine to keep running, and predictable behaviour under load. Same credit balance as the dashboard.
API access is granted on approval. Request it from the dashboard’s API Access page — ready-to-use integration examples are published in the NumberChecker GitHub examples repository.
Pick the pattern that matches when a bad number actually causes you a problem.
Check the moment a user submits a signup or lead form. Blocks fake and mistyped numbers before they enter your database. Keep it asynchronous so latency never blocks the response.
Confirm the number can receive order and delivery notifications before you promise to send them. Cheaper than a failed delivery and a support ticket.
Batch every contact added or edited that day and write the result back as a custom field. Keeps a large database clean without a manual export.
Validate the target segment in one bulk call before a send. Highest throughput and lowest cost per record, with nothing user-facing.
Illustrative only. Build against the documentation in your dashboard once access is approved.
# Submit a batch of numbers for WhatsApp validation curl -X POST "https://api.numberchecker.cloud/v1/whatsapp/check" \ -H "Authorization: Bearer $NUMBERCHECKER_API_KEY" \ -H "Content-Type: application/json" \ -d '{"numbers": ["+447400000123", "+4915100008842"]}' # Response: one labelled record per submitted number { "results": [ { "number": "+447400000123", "valid_whatsapp": true, "country": "GB" }, { "number": "+4915100008842", "valid_whatsapp": false, "country": "DE" } ] }
Confirm the live endpoint, auth scheme, batch limits and rate limits against your NumberChecker dashboard before writing production code.
Most integration pain comes from these decisions rather than from the API itself.
Convert to E.164 with the country dial code, or you will pay for checks on unparseable input.
Treat a result as fresh for 60 to 90 days rather than re-checking on every campaign.
Queue validation and update the record asynchronously so nothing waits on the network.
A batch response can contain records that could not be evaluated. Retry those separately.
Keep country and dial code, not just a boolean — they are what make segmentation possible later.
A 250,000-contact CRM costs about $6.25 to validate. Re-running it weekly without caching costs 52x for no new information.
Validate in real time when a bad number blocks something immediately — signup, checkout, or any flow where you are about to promise a notification you cannot deliver.
Validate in batches when you are cleaning an existing database or preparing a campaign list. Batching is cheaper per record, avoids adding latency to user-facing requests, and lets you retry failures without a user waiting.
Most teams end up doing both: real-time at the point of capture, plus a scheduled sweep to catch decay in records captured months ago.
At $1 per 40,000 checks, the cost of validation is rarely what constrains an integration — repeated unnecessary checks are.
Cache aggressively with a timestamp and re-validate on a schedule instead. Credits do not expire, so there is no penalty for buying ahead of a large migration.
WhatsApp registration is one signal among several. If you also need line type, carrier, or whether a number is VoIP rather than mobile, those are separate lookups available in the same platform.
Deciding which signals you genuinely act on — before you start storing all of them — keeps both cost and schema complexity down. A field nobody queries is a field that will drift out of date silently.
Yes — a REST API is available from NumberChecker.cloud on approval, sharing the same credit balance as the dashboard at $1 per 40,000 checks.
No. Validation happens server-side with no login, QR scan or linked device.
E.164 — full international format with the country dial code and no spaces or punctuation. Normalise before sending so you never pay for a check on unparseable input.
Store it with a timestamp and treat it as fresh for 60 to 90 days. That avoids paying repeatedly for the same number while still accounting for contact data decay.
Credits are shared between the dashboard and the API, so you can validate a sample by hand first and then point your code at the same balance.
