WhatsApp API pricing compared: per conversation, per message, per number, per server
The four pricing structures, where each one wins, and an honest decision table — including the volumes where WALayer is the wrong choice.
Every WhatsApp API provider prices differently, and the differences are structural, not cosmetic: per conversation, per message, per number, or per server. Which structure wins depends almost entirely on your volume and traffic shape, so this post compares the structures honestly — including the cases where ours loses.
The four pricing structures
Per conversation (Meta's official Cloud API). You pay for 24-hour conversation windows, at rates that vary by country and category. Marketing conversations to India, Brazil or Indonesia cost multiples of what utility conversations to the US cost. Service conversations — replies to people who message you first — are free. The bill moves with traffic, with market mix, and with Meta's periodic rate revisions. We keep a full per-market breakdown with the current rates in the cost calculator.
Per message (metered gateways). Some unofficial gateways meter per message, typically fractions of a cent. Cheap at low volume; at sustained volume the meter is the product, and a campaign that performs well costs more than one that flops.
Per number, flat (WALayer and most hosted gateways). You pay for each connected number and messaging is not metered. The bill is the same in a quiet month and a peak one. The structural consequence: effective cost per message falls the more you send, instead of rising.
Per server (self-hosted open source). Evolution API, WAHA and similar projects cost nothing to license and everything to operate: a VPS, storage, upgrades every time WhatsApp changes the protocol, and an on-call human when sessions corrupt at 2 a.m. Honest accounting puts a competent self-host at a few hours of engineering a month — which is either trivial or the most expensive option on this page, depending on what your hours cost.
Where the flat-rate structures differ from each other
Among per-number providers, entry pricing in this category starts around $6 per number — and our Solo tier now matches it exactly. Above the entry point we are still not always the cheapest sticker price, and we do not claim to be.
What the sticker comparison hides is what is metered behind it. Some providers cap messages per tier and sell the cap back as an upgrade; some meter media or group messages separately; some quote a per-number price that assumes annual commitment. Our structure is: unlimited, fair-use-metered messaging on every tier (no per-message fees, throttled by pace rather than capped by volume, media and groups included), extra numbers as prorated add-ons, and 20% off for annual — with the delivery layer, warmup automation and ban handling as the actual argument for the premium.
The honest decision table
Choose the official Cloud API when your volume is low, your traffic is mostly inbound support (service conversations are free), or compliance requires Meta's sanctioned channel. At 2,000 messages a month to a cheap market, the official API costs a few dollars. No flat-rate plan beats that.
Choose a metered gateway when you send occasionally and unpredictably — hundreds of messages a month, not thousands — and the number is not business-critical.
Choose flat-rate when volume is steady and meaningful, recipients are in expensive markets, or traffic is proactive: notifications, confirmations, alerts, campaigns. A Sri Lankan business sending 25,000 marketing messages a month is choosing between roughly $1,297 on official rates and $6–45 flat. That is not a close call, and the same shape holds in Germany, Brazil and most of the Gulf.
Choose self-hosted when you have the engineering appetite and the session-ownership problem is interesting to you rather than terrifying. We wrote up why sessions drop if you want to know what you are signing up to operate.
Run your own numbers
The cost calculator computes the official-API bill for your market and volume mix against our ladder, add-ons included, and tells you when we are the wrong choice. That is not false modesty; a customer who picked flat-rate at the wrong volume churns in a month, and we would rather lose the signup than the trust.