The headline price is not the price.
Four billing models dominate this market, and they produce wildly different bills for the same business. Work out which one you are being quoted before you compare any two numbers.
Charged for capacity
Per user seat, or per contact reachable. You pay for the size of the thing.
Charged for usage
Per conversation, per reply, or metered credits. You pay for what happened.
What each model does to your bill
| What each model does to your bill | Charged for capacity | Charged for usage |
|---|---|---|
| Per user / seat | Every staff member who needs the inbox adds a fixed monthly cost. | Punishes growing the team. A five-person shop can pay more for seats than for the product. |
| Per contact reachable | Priced on how many people you could message this month. | Your bill rises with your contact list even in a quiet month, and old contacts you never message still count. |
| Per conversation or reply | A fixed amount each time a conversation opens or a reply is sent. | Simple to picture, but a long conversation and a one-line answer cost the same. |
| Metered usage / credits | Priced on the work actually done answering. | A quiet month is genuinely cheap. Needs a clear meter, or it is impossible to predict. |
| Meta fees | On the Cloud API these are extra, whatever the model above. | Ask whether they are passed through at cost or marked up. Both happen. |
| Extra channels | Telegram or a website widget is often a separate line, or a higher tier. | Ask before you buy, not after. |
| Extra WhatsApp numbers | Very often another full subscription each. | This is the one that surprises people. Some plans cap you at one number at every price. |
How to work out your real cost
Count the things that will grow
Staff, contacts, numbers, channels. Whichever of those your vendor charges for is the one that will make next year expensive. A price that is fine for one person and one number can multiply quietly.
Ask what a quiet month costs
Seasonal businesses pay capacity pricing all year for three busy months. Usage pricing follows the season instead. If your trade has an off-season, this is the single biggest question on the page.
Where aiBalas lands
We charge per agent, per month, and meter the answering in RM — so a quiet month is cheap and a busy one scales with the value it produced. What we deliberately do not charge for is the things that usually grow: an agent answers on WhatsApp, Telegram and a website widget at once for one price, holds up to five WhatsApp numbers at no extra cost, and adding your staff to the workspace costs nothing. There is a free tier with a monthly RM allowance included, so you can see your own real usage before deciding anything.
Common questions
Why do so many vendors charge per contact?
Because it grows automatically, which is good for the vendor and bad for you. It also matches how Meta used to bill, before it moved to per-message pricing — so some of it is inherited rather than designed.
Are Meta's fees included in a subscription?
Almost never, and the good vendors say so plainly on the pricing page. On the Cloud API you are paying two parties: the software vendor and Meta. Ask whether Meta's rate is passed through at cost or with a margin on top.
Is metered usage harder to predict?
It is only unpredictable if the meter is vague. Ask what a typical conversation costs and whether you can see usage as it accrues. If a vendor cannot answer that in one sentence, that is the real answer.
What should I actually compare between two quotes?
Build the same scenario for both: your number of staff, your contact list, your busiest month, every channel you want, and every WhatsApp number you need. Compare those two totals. Headline prices in this market are not comparable.
Try it before you decide
Free to start, no credit card. WhatsApp, Telegram and website chat on one agent.