Most hiring software rents you access and wishes you luck. Metix AI meters the thing you were trying to buy in the first place: an interested, hire-ready candidate, reviewed by our delivery team and booked on your calendar. A seat bills whether or not anyone arrives. A credit moves only when a qualified candidate does.
Metix AI prices hiring by the delivered candidate, not the seat. A credit is one interested, hire-ready candidate, checked by our delivery team before delivery and booked on your calendar. If nobody clears your bar, no credit is spent. Plans run $49 to $299 a month with 3 to 75 credits, annual billing saves 10 percent, and every plan starts with a free 14-day trial: 3 New Roles, 12 Credits, no card required.
Every pricing model names a unit, and the unit is the most honest sentence a vendor writes. Seats. Contact reveals. A percentage of first-year salary. Whatever the marketing says, the unit is what the company gets paid to produce more of. Ours is a person. One credit unlocks one interested, hire-ready candidate, and if they're not worth meeting, you don't spend it.
That definition has three working parts. Interested means the candidate has said yes to your specific role in a real conversation, not that a database flagged them open to work. Hire-ready means they match the role's must-haves and arrive ready to talk about the job rather than the basics. Reviewed means a person on our delivery team has checked them against your bar before anything reaches you. Behind the definition sits the pipeline: Mira, our AI recruiter, turns your role brief into a search, sources across 860M+ profiles in 190+ countries, sends outreach you approve first, and screens for genuine interest and fit. The system behind her is published research, not a black box. A credit buys the far end of all that work: the person, delivered.
The list of things a credit is never spent on matters just as much. Not a search. Not a profile view or a contact reveal. Not outreach, however many messages a role takes. Not a promising profile who never answers. Those are our costs. Pricing them into the unit, instead of billing them through to you, is the whole idea.
Every pipeline contains a fraction that looks right on paper and is wrong in the room. Screening exists to shrink that fraction, and screening quality is therefore a risk someone has to hold. The real question in any hiring contract is who.
Under a seat license, you do. The vendor is paid in advance, in full, whether the month produced twelve strong candidates or none. Their most profitable customer is the one who pays and barely logs in, a flaw we unpacked at length in our essay on outcome pricing. Nothing in that model punishes a sloppy screen. You feel the miss in interview hours. The vendor never feels it at all.
A credit moves the risk across the table. We spend real effort before your meter does: the sourcing, the outreach, the screening, and a delivery lead's review are all sunk into a candidate before you decide whether they clear the bar. Someone our review rejects costs us and costs you nothing. That changes behavior on both sides. We have no reason to pad a shortlist, because padding is priced. You have no reason to ration your attention, because looking is free.
The contrast is sharpest in a slow month. A seat bills the same in a month with five hires and a month with none. That is the design of rented capability, and for a staffed sourcing team with hours to spend, it can be the right deal. We walked the seat-versus-candidate trade in detail in Metix AI vs hireEZ, which sets hireEZ's published seat pricing against ours. The units differ, and the units are the argument.
The rule fits in one line: a credit only unlocks a candidate who is qualified and genuinely interested, so if no one clears the bar, nothing is charged. A candidate who misses your bar never reaches you and never costs a credit. There is no partial charge for effort, and no fee for a shortlist you didn't ask for.
One honest footnote belongs next to that rule. The plan itself bills monthly, like any subscription. What the rule governs is the credit meter inside it: a month that delivers nobody consumes no credits, so your invoice maps to people you actually met rather than to time elapsed.
Read as an incentive, this one rule does most of the work on this page. Every delivered candidate carries our sunk cost, so the cheapest way for us to operate is to screen well the first time. A vendor paid on access can afford a noisy funnel. We can't.
The plans put numbers on the meter. Plans run $49 to $299 a month: Starter at $49 with 3 credits, Growth at $89 with 12, Professional at $159 with 30, and Scale at $299 with 75. Active roles scale alongside, from 1 to 12. Annual billing saves 10 percent. Plans are monthly with no long contract, you can cancel anytime, and every plan starts with a free 14-day trial: 3 New Roles and 12 Credits, no card required.
| Starter | Growth | Professional | Scale | |
|---|---|---|---|---|
| Price per month | $49 | $89 | $159 | $299 |
| Credits per month | 3 | 12 | 30 | 75 |
| Active roles | 1 | 3 | 6 | 12 |
Monthly list prices from the pricing page. Annual billing saves 10 percent on every plan.
The steepest comparison is not with software at all. A headhunter typically charges 20 to 30 percent of first-year salary. Retained search bills that whether or not the role is filled; contingency search bills on placement, which is better, and still lands the whole fee in one invoice. A single senior placement fee can cost more than a full year on a Metix AI plan. To be fair, the agency does finish the search, and finishing is exactly what we price too. The difference is the denominator.
Speed is part of what the credit buys, because a slow qualified candidate is often a lost one. From the early roles Metix AI has run, and moving with the role and the market:
Outcome pricing has a known failure mode: it is only as honest as its metric. Pay a support vendor per resolved ticket and tickets get closed that reopen the next morning. Hiring has its own versions, they are predictable, and we would rather name them than hope you never ask.
Move one: loosen the definition. A careless vendor stretches qualified until the inbox fills with people who pass on paper and fail in practice. Here the definition is not ours to stretch. You set the bar in the role brief, and you approve every outreach message that goes out under it, so the standard each candidate is sold on is one you signed off.
Move two: flood the funnel. Volume can imitate quality for a while. It fails against the review step. A delivery lead checks every candidate before delivery, and anyone who misses the bar costs no credit, so flooding spends our review time without ever touching your meter.
Move three: creaming, the oldest trick in pay-for-results. Pour effort into the easy cases and park the hard ones. It has nowhere to hide here, because you point Metix AI at the role you choose. There is no portfolio of easy reqs for us to retreat to. If we cream, you see it on the one role you brought, which is why we keep saying: bring the hard one first. The canonical wording of what a credit unlocks lives in the FAQ, and the longer argument about gamed metrics is in the essay linked above.
The interview is still yours, and so is the decision. We stop at a qualified, interested person booked on your calendar. Nobody here pretends to make the hire for you: you run the interview, you sign the offer, you employ the person, and you own the relationship from there.
It also does not make Metix AI the right buy for everyone. A staffed sourcing team that wants an engine to run its own funnel is often better served renting one, and we say so in the comparison above. Outcome pricing is for the buyer who wants the output itself: a meeting with someone worth meeting, at a price that moves only when that meeting exists.
The cleanest test of a pricing model is a live role, ideally your hard one. The trial is 14 days with 3 New Roles and 12 Credits, no card required. You define the bar, you approve the outreach, and the first credit moves only when someone worth meeting lands on your calendar. Then read the invoice and check what it maps to.
One credit unlocks one interested, hire-ready candidate: a person who matches the role's must-haves, has confirmed interest in a real conversation, has been reviewed by the Metix AI delivery team, and is booked on your calendar. Credits are never spent on searches, profile views, contact reveals, or outreach.
Nothing is charged. A candidate who misses the bar never reaches you and never costs a credit, so a month that delivers nobody consumes no credits. The plan itself bills monthly like any subscription; the no-charge rule governs the credit meter inside it.
Plans run $49 to $299 a month: Starter at $49 with 3 credits, Growth at $89 with 12, Professional at $159 with 30, and Scale at $299 with 75. Annual billing saves 10 percent, and every plan starts with a free 14-day trial with 3 New Roles and 12 Credits, no card required.
You set the bar in the role brief and approve every outreach message that goes out under it, a delivery lead reviews every candidate before delivery, and anyone who misses the bar costs no credit. Creaming has nowhere to hide because you point Metix AI at the role you choose.