Skip to content
Menu
Home / Vendor profile

ZeroBounce

We tested ZeroBounce with fake control addresses and re-checks months later. Good addresses failed at about 1.35% a month, so we re-verify on a schedule.

Product typeEmail verification
Pricing$39 · one-time credit purchase
Official sourcezerobounce.net
Visit website ↗
Our testJan to Nov 2026 · about 16,000 cached verdicts · 200-address re-check in Aug 2026All our tests ↗
The short answer

ZeroBounce did what a send gate needs in our 2026 use, as far as we could test without sending to everything. A made-up control address told us when its valid verdicts meant something, and re-checking 200 good addresses showed them failing at about 1.35% a month, so we expire verdicts after 90 days.

ZeroBounce was the send gate on every outbound list we built from January to November 2026, and by November our cache held about 16,000 verdicts. The question for a review is whether you can trust what it tells you. We didn't send to every address it checked, so we have no ground truth for its accuracy. What we could do was test its verdicts in two ways that don't depend on sending: a made-up control address on the same domain, and re-checking addresses months after they first passed.

A control address shows when "valid" means something

A catch-all domain accepts mail for any address, so an SMTP check there can't tell a real mailbox from an invented one. The verdict on a guess is only worth paying for if the domain would reject a fake.

So we added a fake. In October 2026 we needed one person's address at a company where we only knew the name. We sent ZeroBounce six format guesses plus a made-up control address at the same domain, 7 credits in all. The mail server rejected the control, which meant the domain checks individual mailboxes, so the guess that came back valid could be trusted. On a second company the same day, the domain was catch-all, and no number of guesses would have settled it.

Our bulk pattern probe followed the same rule: as soon as a domain answered catch-all, the script stopped spending there and flagged the company unverified. The test is cheap enough to run before any guessing, and the alternative is paying a credit per guess for the same non-answer.

Good addresses go bad at about 1.35% a month

In August 2026 we re-verified 200 addresses ZeroBounce had already rated valid, sampled with extra weight on the oldest verdicts so the aging window was as long as the cache allowed. 190 were still valid, 5 had become invalid, and 5 had moved to catch-all or unknown. All five invalids came back mailbox_not_found: the server answered and said that one mailbox no longer existed, which is what a departure looks like, as opposed to a dead domain.

Weighted by how long each address had been in the cache (370.5 address-months in total), that is 1.35% of good addresses failing per month, with a 95% interval of 0.44% to 3.15%. All five failures came from the oldest group, about 2.3 months old, which failed at 2.15% a month.

That pattern points to a lag. Companies tend to keep a leaver's mailbox open for a while, so a young address can't fail yet even if the person has gone. We read 1.35% as a floor on decay, and it's why we kept a 90-day expiry on cached verdicts.

Catch-all limits re-checking too. Across the cache, 4.1% of records sat on catch-all domains, where a re-check can't see anyone leave.

How we used the six statuses

ZeroBounce status What our send gate did
valid Sent
catch-all Held from cold email, allowed for ad audience matching
unknown Held from cold email, allowed for ad audience matching
invalid Suppressed
do_not_mail Suppressed
abuse Suppressed

About 88% of our cache was valid, and about one verdict in eighteen was catch-all or unknown. Keeping those two apart from valid cost us some reach and kept unconfirmed addresses away from our sending domains. ZeroBounce also checked the revealed addresses in our QuickEnrich review, and it was the last stage of the recipe on what is waterfall enrichment.

Two calls you need

GET https://api.zerobounce.net/v2/getcredits?api_key=YOUR_KEY
GET https://api.zerobounce.net/v2/validate?api_key=YOUR_KEY&email=name@example.com&ip_address=

Read the balance before a batch and stop if it's lower than the batch needs. Log the HTTP status and error body for each call, or an out-of-credits run will look like a list full of bad addresses.

Pricing

Dated pricing record$39 · one-time credit purchase

Confirm the selected configuration and usage allowance.

In November 2026 the smallest pay-as-you-go purchase was $39 for 2,000 validations. Our credits didn't expire, which suited usage that came in bursts of a few hundred. More on ZeroBounce pricing.

Who it's for

Worth it if Skip it if
You send cold email and need a gate before each send You only mail opted-in subscribers
You build addresses from patterns and need each guess checked You need answers on catch-all domains, which no SMTP check gives
Your usage comes in bursts and you want credits that last Your tools collapse every result into valid or invalid

Run both checks yourself

  1. Before guessing at a domain, verify one made-up address there. If it comes back valid or catch-all, stop.
  2. Take 200 addresses you verified 60 to 90 days ago, re-check them, and divide failures by total address-months.
  3. Split the result by age. If only the oldest addresses fail, your real decay rate is higher than the average.
  4. Store each verdict with its date and set an expiry before you scale up.

Questions buyers ask

Is ZeroBounce accurate?

We can't give an accuracy rate, because we didn't send to every address to get ground truth. Its verdicts passed the checks we could run: a made-up control address was rejected where real guesses passed, and re-checks months later showed good addresses failing at about 1.35% a month.

How often should you re-verify email addresses?

We re-verify after 90 days. When we re-checked 200 addresses rated valid months earlier they had failed at 1.35% a month, and the oldest group at 2.15%, so treat the average as a floor.

How do you verify emails on a catch-all domain?

You can't settle them with an SMTP check. Verify one made-up address at the domain first; if it comes back valid or catch-all, stop spending credits there.

What should you do with catch-all and unknown results?

Hold them out of cold email. We kept them for ad audience matching only and sent only to valid.

Do ZeroBounce credits expire?

Our pay-as-you-go credits didn't. In November 2026 the smallest pack was $39 for 2,000 validations.

Sources

Compare alternatives

Compare the part of the workflow you want to replace, including scope, billing and operating responsibility.

ZeroBounce alternatives →
Sources and evidence behind this page