Should I use my own ZeroBounce API key in Clay?

At any real volume, yes. Clay's managed ZeroBounce account costs one Clay credit per enriched cell. Supplying your own key spends ZeroBounce credits instead, which is materially cheaper. Bring-your-own-key is available on Clay's paid plans and is a settings change rather than a project.

How accurate is ZeroBounce really?

ZeroBounce's own Trustpilot profile gives a segmented figure: over 96 percent on AOL, Yahoo and Office 365, and 99 percent elsewhere. Independent benchmarks put real-world accuracy at 96 to 98 percent. We use the segmented numbers rather than the flat marketing figure, because independent testing does not reproduce the latter.

What does a catch-all result mean?

The domain accepts mail for any address, so SMTP verification cannot determine whether that specific inbox exists. It is a legitimate third category, not an error. Treating catch-alls as invalid discards reachable contacts; treating them as valid risks deliverability. Clay's Safe to Send toggle excludes them, which is the conservative default.

Does ZeroBounce charge for unknown results?

No. ZeroBounce documents that the API never consumes a credit for an unknown result, so greylisted or timed-out validations are free. Verification is otherwise one credit per email, identical for real-time API and bulk, and AI scoring is also one credit per address.

Are there limits on bulk validation?

ZeroBounce documents no restriction on file size, number of emails, or number of files submitted through the bulk endpoint, accepting CSV and TXT. The practical constraint is your credit balance. Per-second rate limits and concurrency limits are not published, so we do not rely on any figure for those.

How do you use ZeroBounce with Clay?

Clay ships a native ZeroBounce integration with a single action, Validate Email, documented by both companies. It works, it is straightforward, and most teams need no help with it. The decision worth making deliberately is whether to use Clay's managed ZeroBounce account or bring your own API key, because that determines which credits you burn.

What already exists

Is there a native ZeroBounce to Clay integration?

This is the most straightforward integration on this list and the weakest case for hiring anyone. Clay exposes one ZeroBounce action, Validate Email, which determines whether an address has a valid inbox. It ships a Safe to Send toggle that restricts results to valid only, and auto-update runs every 24 hours by default. Both Clay and ZeroBounce document the integration from their own side, which is a good sign for its stability.

One action only. Validate Email is the whole surface, which is appropriate for what ZeroBounce does.
Catch-all addresses cannot be resolved to valid or invalid by SMTP. This is an industry-wide constraint rather than a ZeroBounce defect, and it is why the Safe to Send toggle and ZeroBounce's separate AI scoring product both exist.
Auto-update is on by default at 24 hours. Turning it off risks stale validation results, which is a quiet way to degrade deliverability over time.
Bring-your-own-key is available on paid Clay plans. Note that some documentation references older Clay plan names that no longer match what Clay sells, so check the pricing page rather than the docs.

The detail

What actually moves between ZeroBounce and Clay?

One action: Validate Email, returning inbox validity for an address

Statuses returned: valid, invalid, catch-all, unknown, spamtrap, abuse, do_not_mail, plus sub-statuses

A Safe to Send toggle restricting results to valid only, excluding catch-alls

Auto-update every 24 hours by default, keeping validation fresh

Conditional runs via formula, so you validate only the rows worth validating

In practice

Where ZeroBounce to Clay syncs go wrong

Paying Clay credits when you could pay ZeroBounce credits

Using Clay's managed ZeroBounce account costs one Clay credit per enriched cell. Supplying your own ZeroBounce API key spends ZeroBounce credits instead, which are materially cheaper at volume. This is not a defect, it is a default, and it is the single highest-value thing to change on a table validating tens of thousands of addresses. Most teams leave the default in place without realising there is a decision to make.

Clay docs: ZeroBounce integration overview, credits and bring your own key

Catch-all results are not a failure, and treating them as one costs you contacts

A catch-all domain accepts mail for any address, so SMTP verification cannot determine whether a specific inbox exists. ZeroBounce reports this honestly as catch-all rather than guessing. Teams that treat catch-all as invalid discard reachable contacts; teams that treat it as valid send into unknown territory and take deliverability damage. The honest answer is that catch-all is a third category needing its own decision, which is exactly why the Safe to Send toggle exists.

ZeroBounce docs: validation statuses, credit model and bulk limits

Validating the same address twice, in two places

A common and avoidable cost: running ZeroBounce inside Clay and then again inside the sequencer, paying for the same verification twice. This is a workflow design issue rather than a product one, and it is worth checking whenever validation costs look higher than expected.

Disabling auto-update to save credits, then sending to stale data

Auto-update refreshes validation every 24 hours by default. Turning it off is a reasonable way to control spend, but email validity decays as people change roles and leave companies, so the saving comes at the cost of accuracy that only shows up later as bounces. If you disable it, do so with a deliberate re-validation cadence rather than never.

ZeroBounce docs: Clay integration

When not to hire anyone

Where the native integration is genuinely fine

Almost always, and we will say so plainly: this is the weakest custom-build case of anything on this page. The native integration handles email validation inside Clay properly, and the main optimisation available to you costs nothing but a settings change, which is supplying your own ZeroBounce API key. A build only makes sense at volumes where Clay's per-row model stops being economic, in which case you batch through ZeroBounce's bulk file endpoint outside Clay entirely, or where you need a catch-all resolution layer chaining a second provider behind ZeroBounce. Neither is a common requirement.

How we build it

What a ZeroBounce to Clay build involves

Usually nothing, and that is the right answer. If a client asks us about this integration, the first thing we check is whether they are on Clay's managed ZeroBounce account, because switching to their own API key is often the entire saving they were looking for.

Where volume genuinely justifies it, we move validation out of Clay and batch through ZeroBounce's bulk endpoint, which has no documented restriction on file size or number of emails, then bring results back. That is cheaper per address but adds a moving part, so we only do it when the arithmetic clearly supports it.

For teams where catch-all volume is high, we build a resolution layer: ZeroBounce first, a second provider or ZeroBounce's own AI scoring behind it for the catch-all subset, and a documented policy for what happens to addresses that stay unresolved.

We check for double validation across the stack, because paying twice for the same verification is common and nobody notices it in an invoice.

Sources

Related integrations

Clay to HubSpot Clay to Salesforce PredictLeads to HubSpot Apollo to HubSpot

If you are weighing whether to build this in house or have it built, read what a build like this costs, or see worked examples from real engagements. The vocabulary behind all of this is in the B2B GTM and automation glossary.

Get in touch

Tell us about your situation

Give us some context and we'll come to the conversation prepared. No generic pitch. No obligation.

We review every inquiry personally and respond within one business day.