Apollo credits explained: what spends them, and how to budget a list

Most people find out how Apollo credits work the same way: they run out. The balance looked fine on Monday, a list got enriched on Tuesday, and by Wednesday the next campaign has nothing to build with.

Credits aren’t complicated, but they don’t work the way the word “balance” suggests. This guide covers the separate pools, what’s free, what each job costs, and a worksheet you can fill in before you build a list. Every number comes from our own paid account, with the date we measured it. Plans and prices change, so check your own usage page before you rely on any of them.

There isn’t one balance, there are eleven

Apollo keeps separate pools of credits, and spending from one doesn’t touch the others. When we read our account on 31 August 2026, it showed eleven. These are the ones that matter for outbound.

The ratio between the first two rows tells you how Apollo thinks about price. We had 200 times more AI credits than lead credits on the same plan. Apollo treats its own AI as close to free and treats data as the thing you pay for. So the expensive part of a list is finding and confirming the people. Writing to them costs almost nothing by comparison.

That’s why “we have 3,000 credits left” doesn’t mean much on its own. Always ask which pool.

What’s free

A lot of the work costs nothing, and knowing which parts are free changes the order you do things in. We measured each of these on our own account.

• Searching for people. We paged through thousands of results and the lead credit balance didn’t move.

• Filtering to verified emails in the search. One search of ours went from 3,240 people to 2,742 with that filter on. That’s about 500 people we never paid to enrich.

• Seeing who has an email before you pay. Search results show whether Apollo has an email for each person. Leaving out the ones without took one list from 2,259 people to 2,019.

• Turning a company name into a record. Looking a company up by name is free. A full company search isn’t.

• Saving people as contacts and adding them to lists or sequences. Once someone is enriched, moving them around costs nothing.

What spends lead credits

Lead credits are the scarce pool, and five jobs draw on it. The table shows what each one cost us.

Two other costs sit in the same pool and catch people out, because they don’t feel like data.

• Buying mailboxes inside Apollo. A shared mailbox cost 300 credits, a Google mailbox 800, and an Outlook mailbox 1,500. Three Google mailboxes is 2,400 credits, which was more than half of our month.

• Waterfall enrichment. When Apollo checks other data providers for you, the cost varies by plan and by result. Nobody can quote you a fixed number in advance, so get the estimate before you run it.

The order that saves the most

Because search is free and enrichment isn’t, the cheapest list is the one where everything free happens first. This is the order we build in.

1. Search narrow, then read what came back. A filter states what you meant, not what you got. One search of ours returned 679 people or 91,649 depending on a single setting. The detail is in Apollo filters that inflate your TAM.

2. Filter to verified emails, and drop people with no email. Both are free, and both remove people you’d otherwise pay for.

3. Remove duplicates. One pull of ours returned 3,238 rows holding 2,933 different people. Without that check we’d have paid for 305 people twice.

4. Remove everyone you shouldn’t email. Customers, open deals, opt-outs, and anyone you’ve already enriched. That last group matters most for credits, because Apollo charges again.

5. Search again if the list has been sitting. A list built on company growth lost 106 of its 1,259 people in 18 days, because they no longer matched. Searching again is free. Enriching the old file would have spent 106 credits on people who no longer fit.

6. Then enrich. By now the number you’re paying for is the number you’ll send to.

A worksheet for one list

Fill this in before you build. The right-hand column is a worked example for a 1,000-person list on a plan with 4,000 lead credits.

On that plan, three lists like this fit in a month with about 970 credits to spare. Buy three Google mailboxes in Apollo in the same month and you’re down to one list.

We work to one more rule. If a job would use more than 85% of the credits you have left, stop and decide on purpose. One of our lists would have cost 2,712 credits with 3,055 remaining. That’s 89%, and it would have left 343, which isn’t enough to do anything else that month. The choice at that point is to enrich only the strongest part of the list, wait for the credits to reset, or go ahead knowing what it leaves you with.

If an agent runs it for you

More teams now build lists through Claude, ChatGPT, or their own agent, using Apollo’s MCP or its command line tool. An agent spends credits much faster than a person clicking, so it needs firmer rules. This section is written for the agent.

• Read the pools before any paid step. apollo usage credits on the command line, or apollo_usage_stats_credit_usage_stats on the MCP. Report the remaining balance by pool name, and never say “credits” without the pool.

• Quote the cost and wait for a yes. State the full scope once, in the wording the tool requires. Don’t confirm batch by batch.

• Apply the 85% rule against the remaining balance, not the monthly limit. Above it, stop and offer three options: enrich the top tier only, skip, or proceed.

• Filter on has_email and on verified email status in the search. Both are free. Records without an email are not worth a credit.

• Dedupe by person id before enriching, and subtract everyone already enriched. Enrichment charges again for the same person.

• Enrich 10 records per call and write each result to disk. A longer file is rejected. Save the full response, because the credits bought it and it lives nowhere else.

• Report credits_consumed from the responses, summed. Don’t report an estimate, and don’t promise that unmatched records are free.

• Stop paging company search at the first empty page. It costs a credit too.

• Never quote a fixed cost for waterfall enrichment. Say that it varies and get that accepted first.

• Re-run the search before enriching a list that has been held back. It’s free, and timing-based lists drift.

When someone else should watch the credits

If you’d rather not think about any of this, there are two ways we can help. Apollo Operator is our free, open-source toolkit, and every rule above is built into it, including the 85% stop. With Apollo Managed, we build the lists inside your account and you see what was spent and why. And if you’re not sure where your credits have been going, the audit and 14-day trial is $1,500 and starts by reading the account.