Pricing in this category is harder to compare than it should be, because vendors bill in incompatible units. One quotes a monthly plan, one quotes per minute, one quotes per order, and one won't quote at all until you take a sales call. None of those numbers can be compared directly.
The way through is to convert everything into a monthly total using your own call data, then add the line items that live outside the headline price. This piece walks through both, and states our own numbers plainly so you have at least one concrete anchor.
Start with your own usage, not a price page
Three numbers, all from your phone provider's call log export:
Total connected minutes per month. Not call count. Minutes is the unit most voice pricing runs on, and call count will mislead you if your calls are long.
Total inbound calls per month. Needed to evaluate per-call pricing models.
Your busiest month in the last year. This is the run that separates pricing models, and it's the one people skip.
Pull ninety days if you can. If your provider makes this hard, most modern phone systems have a usage or analytics export somewhere; it's worth the twenty minutes to find.
What X1 Voice costs
So there's a concrete number in front of you: Starter is $250/month and includes 750 minutes; Professional is $750/month and includes 2,500 minutes. Minutes past the included pool bill at $0.35 on Starter and $0.33 on Professional. There's no setup fee and no long-term contract on Starter or Professional. Enterprise is custom for multi-location groups.
Every feature is on every plan, so the difference between tiers is minutes and overage rate rather than a gated feature list. That's a deliberate choice on our part, and it matters for comparison because it means you don't have to check whether the thing you need is on the tier you can afford.
The line items that live outside the headline price
When you're comparing any vendor, ask for these specifically. Each can be legitimate; none should surprise you in month two.
Setup or onboarding fees. Common in the category. Ask what they cover.
Menu import and menu change fees. Some vendors charge for the initial build and some charge again for significant changes. If your menu changes seasonally, that's a recurring cost.
Per-location fees. Crucial if you might open a second store. Ask whether minutes pool across locations or sit siloed per store, and whether adding a location restarts your contract term.
Telephony and number charges. Porting fees, monthly number rental, and per-minute carrier charges passed through.
SMS charges. If the system texts order confirmations, someone pays per message. Ask whether it's included. See SMS order confirmations.
POS integration fees. Occasionally billed separately, occasionally billed by the POS vendor rather than the voice vendor.
Support tiers. Whether reasonable support is included or costs extra.
Renewal pricing. A first-year rate that steps up is a normal term, but it belongs in your comparison. Watch for uncapped "then-current rates" language, which we cover in contract terms to avoid.
How pricing models behave differently at volume
The four common models are per-minute, per-call, per-order, and flat monthly with an included allotment. We break each one down in AI phone answering pricing models, but the cost-planning summary is short.
Usage-based models track your volume, which means they're gentler when you're small or seasonal and more expensive on exactly the nights you're busiest. Per-order pricing in particular gives your bill the same shape as delivery commissions, which is the cost structure many restaurants adopt phone AI to avoid — see voice AI vs. delivery app commissions.
Flat models are predictable and easier to budget, and they mean a busy December doesn't rewrite your P&L. The tradeoff is that a genuinely quiet location pays for headroom it doesn't use.
Run your normal month and your busiest month through every vendor's rate card. If a model looks close in October and diverges in December, you've found the thing worth knowing.
Building the comparison table
For each vendor, produce one number: expected total monthly cost in a normal month, and one for your busiest month. Include every line item above, amortizing any one-time setup fee over twelve months so it appears in the monthly figure.
Then, separately, note what you'd pay to leave: contract remainder, data export fees, porting costs. Exit cost is part of total cost and almost nobody prices it.
That's a comparison you can defend to a partner or a lender, which a stack of pricing pages is not.
The other side of the ledger
Cost only means something against what it returns. The return side has three components, and we'd rather you build it from your own records than from our examples: recovered orders you currently miss, labor time you stop spending on the phone, and any delivery commission avoided by keeping callers direct.
The honest caveat is that the whole model hinges on your miss rate, which is why measuring it matters more than any price comparison.
If you barely miss any calls, the return side is thin regardless of price, and the right answer may be to spend the money elsewhere. We'd rather say that than pretend every restaurant has a phone problem.
What to expect if you're small
For a single location with modest call volume, the realistic picture is a few hundred dollars a month against a return that depends almost entirely on how many peak-hour calls currently ring out. That question gets its own treatment in is AI phone answering worth it for a small restaurant, including the cases where the answer is no.
The bottom line
Get your own minute and call counts first, then convert every vendor's pricing into a monthly total for a normal month and your busiest month, including setup, per-location, SMS, telephony, and renewal terms. Compare those totals, not the headline rates. Ours is $250/month at Starter and $750 at Professional with no setup fee and no contract, which is a floor for the arithmetic rather than an argument. What determines whether any of it is worth paying is on the other side of the ledger, and that number is yours to measure.