ATS pricing in 2026 has six distinct models, and vendors mix them freely enough that two quotes for "an ATS with AI" can differ by 100x. This guide decodes each model with real numbers (official pricing pages as of July 2026; reported figures marked), names the trap hidden in each, and ends with our obviously-interested-but-checkable view of what screening should cost. (Disclosure: Rubrily is our product. Prices change — verify before buying. Refreshed quarterly.)
The six models
1. Flat company-wide subscription. One price, everyone in. Breezy HR runs $157–$439/mo (annual billing); Ashby's Foundations is $400/mo up to 100 employees. The trap: "flat" usually has a headcount asterisk — see model 3.
2. Per-recruiter seat. Manatal anchors the market's floor at $15–$55/user/mo; Zoho Recruit runs ~$25–$75/recruiter (reported); agency-grade tools like Loxo jump to $169/user. The trap: per-seat punishes collaboration — every hiring manager who wants real access becomes a line item, so teams share logins and the audit trail dies.
3. Employee-headcount bands. Workable's $299/$599/$719 monthly tiers apply to companies of 1–20 employees; bigger bands pay more, and reviewers report sharp jumps at thresholds (reported: crossing 20 employees can nearly double the bill). Ashby, Recruitee, and Teamtailor scale the same way. The trap: you pay for company size, not hiring activity — a 200-person company that hires twice a year subsidizes the vendor's 200-person logo that hires weekly.
4. Per-job or job-slot pricing. JazzHR's entry plan caps monthly postings; Zoho meters active jobs per license; Homerun sells slots. The trap: job caps convert a hiring push into an upgrade event — the pricing binds exactly when you need the tool most.
5. Usage-metered AI — the 2026 pattern. This is the year's structural change: AI screening priced per unit. Workable's Agent bills ~$0.095–$0.12 per candidate processed after 1,000 free credits; Breezy sells AI evaluation credits ($30 per pack); standalone AI-interview tools reportedly run ~$4 per interview at SMB scale, while enterprise interview platforms price annual contracts from five figures (HireVue's reported ranges start around $25k–40k/yr). The trap is the whole model: your cost peaks precisely when a role goes viral — 3,000 applicants is both the moment AI screening is most valuable and the moment the meter spins hardest. Usage pricing makes thorough screening a budgeting decision.
6. Quote-only enterprise. Greenhouse, SmartRecruiters, HireVue, Lever: no public prices; reported medians run ~$12k/yr (Greenhouse) to $50k+ (HireVue), plus implementation fees and multi-year terms. The trap: opacity is the point — you can't comparison-shop what you can't see, and renewals reportedly drift up 8–15%.
What the market actually pays
The most-cited industry benchmark (SelectSoftwareReviews' pricing guide, updated Jan 2026 — third-party) puts typical annual ATS spend at $250–$3,000 for sub-100-employee companies, $3k–15k for 100–500, and $125k+ at enterprise. Sanity-check against list prices above and the bands hold. Two costs those benchmarks exclude: add-on creep (texting, video, assessments sold separately can add 30–50% to sticker) and the big one — human screening time, the 50–200 hours per role that most ATS pricing leaves entirely on your calendar.
How much should an AI ATS cost?
For the software: small teams shouldn't pay more than low-thousands per year, and genuinely free tiers now exist (compared here). For the AI: be suspicious of paying per candidate screened — it prices the tool's core value as a penalty on your success. The right question isn't "what's the monthly fee" but "what does screening 1,000 applicants cost me, in dollars and in hours?" If you're still building your shortlist of vendors, we compared the whole AI-first field — pricing included — in the best AI ATS tools in 2026.
Where we stand (and how we make money instead)
We built Rubrily's pricing as a deliberate inversion of model 5: screening is the free part — unlimited candidates, unlimited AI interviews, unlimited CV evaluations, for every company, no credit card. Paid tiers (Growth, Scale) sell workflow automation, API/SSO, and support — never screening volume. The business is funded by optional candidate-side add-ons (like premium interview practice), charged only if a candidate chooses one — and recruiters can disable those offers entirely, in which case candidates never see them.
Judge that model as skeptically as the other five. Its honest trade-offs: our revenue depends on candidate choice rather than your invoice, and we don't bundle an HRIS — if you want payroll and performance in the same vendor, a suite like Workable is the better shape for you. What the model buys you is structural: the viral job post, the thousand-applicant pile, the decision to screen everyone instead of a sample — none of it costs you anything, which means thoroughness stops being a line item. That's the property we'd tell you to shop for even if you don't pick us.
A buyer's checklist
Model your true year-one cost with four numbers: expected applicant volume across all roles (drives credit/usage models), employee headcount at year end (drives bands — price the next band, not the current one), seats that genuinely need access (drives per-seat), and add-ons you'll actually enable. Then ask every vendor the same two questions: what happens to my bill if applications triple, and can I see why any AI score was given? The first exposes the meter; the second exposes the black box. Good answers to both are rarer than they should be.
FAQ
Why is AI screening usually an extra cost? Because compute costs vendors real money per candidate, and metering passes that through with margin. It's defensible economics and bad incentive design — it taxes thoroughness. Alternatives exist: bundling AI into flat tiers, or (our model) funding it from optional candidate-side purchases instead of recruiter-side meters.
Are quote-only prices negotiable? Almost always — reported procurement data consistently shows negotiated figures well below initial quotes, multi-year discounts, and renewal increases that can be capped if you ask. Never accept a first enterprise quote; never sign multi-year without a renewal cap.
What hidden costs should I watch for? Implementation fees (five figures at enterprise), per-band jumps at headcount thresholds, add-on unbundling (texting/video/assessments), usage overages on AI credits or interviews, and short data retention on free tiers that forces upgrades to keep your own candidate history.
Screening shouldn't have a meter. Unlimited AI interviews and CV evaluations, free — see the whole model, add-ons and all, on one page. See pricing →