The first API bill can be intimidating for those who don’t understand the billing logic — and reassuring for those who do. The cost is not per contact or per message exchanged: it depends on who initiated the conversation and which category the message falls into. This article explains the calculation and shows how to plan for it.
In this article
The billing logic
Two rules explain almost everything. First: who initiates the conversation determines if there is a charge. Messages sent by the customer to you, and your replies within the service window, do not generate costs — the platform encourages companies to provide service.
Second: when the company initiates, the price depends on the category of the template sent. Categories are marketing (promotion, offer, invitation), utility (confirmation, order update, appointment reminder, invoice collection), and authentication (verification code). Marketing is the most expensive; utility is significantly cheaper; authentication has its own price.
It’s worth noting an important model change: billing started based on conversations lasting 24 hours and evolved to billing per message of the template sent, with conversations initiated by the customer no longer charged. If you read older material mentioning "cost per conversation," this is what changed.
Since pricing tables vary by country and are updated periodically, memorizing values doesn’t make sense: what matters is understanding the structure and consulting the current official table for Brazil before planning campaigns. What doesn’t change is the logic — and that’s what allows you to reduce the bill.
What is free (and most don’t use)
Three situations do not generate message costs and together cover a large part of a service operation’s volume:
Customer-initiated conversations. Someone writes to you and the team responds. All exchanges within the service window fall here. For companies whose volume comes from website, ads, and referrals, this means most service is free.
Utility messages within the open window. When the conversation is already active because the customer wrote, transactional notices tend not to be charged. This reinforces the previous point: keeping the conversation alive is economically advantageous.
Free entry points. Conversations originating from ads that open WhatsApp and from the button on business pages usually have an extended service window with no message charges within it. Those investing in conversation ads should know this detail: it changes the campaign’s cost.
The strategic consequence is clear: operations designed so that the customer starts the conversation spend much less than operations designed to send messages proactively. This is not just savings — it also preserves the number’s reputation.
Continue reading for free
Fill out once and unlock this and all other blog articles. We charge nothing — we just want to know who we’re writing for.
Takes 20 seconds. Prefer to talk directly on WhatsApp?
How to estimate the monthly bill
Build the estimate in four lines. Use values from the current official table instead of examples — here the method matters.
1. Customer-initiated conversations. Add up everything coming from website, ads, Google, referrals, and returning customers. Message cost: zero. In many operations, this is 70% or more of the volume.
2. Utility messages you initiate. Appointment confirmations, order notices, reminders, and collections. Multiply quantity by utility value. Example with hypothetical value of R$ 0.08: 900 messages = R$ 72.
3. Marketing messages you initiate. Campaigns, reactivation, promotions. Multiply by marketing value. Example with hypothetical value of R$ 0.40: 600 messages = R$ 240.
4. Authentication, if you use it. Verification codes, with their own value.
In the example above, Meta’s bill would be around R$ 312 for the month — and the total operation cost also includes the platform subscription you use to access the API, detailed in how much a WhatsApp CRM costs.
Do this calculation before approving a campaign. An action for 5,000 contacts in the marketing category has a predictable cost, and it must be included in the return calculation along with the expected response rate.
Five ways to reduce the bill
None involve sending less to those who matter. All involve sending better.
1. Correctly classify the template. An appointment notice is utility; if you insert an offer in it, it becomes marketing and changes price. Keeping transactional messages free of promotion reduces cost and improves readability — the category criteria are in how to write and approve templates.
2. Have the customer initiate. Button on the website, link in ads, QR on printed material, calls on social media. Each conversation started by them is service with no message cost — and with a much higher response rate.
3. Resolve within the window. Conversations dragging on for days require reopening by template. Responding quickly and concluding the matter in the same cycle saves money and sells more.
4. Segment campaigns. Sending to the entire base multiplies cost and reduces response rate. A smaller, targeted list costs less and converts more, as in the group reactivation method.
5. Clean the base. Invalid, duplicate, or opted-out numbers waste money with no return — and increase the risk of complaints.
How to plan a campaign with cost per message
With per-message billing, campaigns stop being just a marketing decision and become a return-on-investment calculation. The reasoning, in four steps:
Define total cost. Number of contacts multiplied by category value.
Estimate response rate based on your history, not supplier promises. Campaigns to recent bases respond much better than to old bases.
Calculate cost per conversation generated. Total cost divided by expected responses. This number is compared to the cost per lead from your paid channels — and is usually much lower.
Set the return floor. How many sales the campaign needs to generate to break even. If the answer is "one," risk is low and decision is easy. If "fifteen," it’s worth reviewing segmentation before launching.
And one discipline to avoid surprises: track accumulated spending weekly, not just at closing. A misconfigured campaign that sends twice is the kind of error that appears on the bill before it shows in the report.
What about QR Code connection users?
Tools that connect to WhatsApp via QR Code do not go through Meta’s billing: there is no cost per message or categories. This makes the bill simpler and, at first glance, cheaper — and that’s exactly why it’s worth understanding what you trade off.
What you save. All message costs. In operations with heavy active sending, this can represent hundreds of reais per month.
What you lose. Formal guarantees: no approved template, no official quality metrics, no tiered limits, and safe volume depends entirely on your behavior. The risk of blocking exists and falls on the number, not the platform.
When each makes sense. Service-centered operations with moderate volume and no need for large-scale sending work well via QR and save money. Operations relying on recurring active communication — confirmations, notices, campaigns — usually prefer the API’s predictability, even paying per message.
An honest reading of the choice: the API charges money for what QR charges risk. Which currency weighs more depends on how much your operation depends on WhatsApp for revenue.
Errors that inflate the bill
Using marketing where utility fits. Inserting an offer in a transactional reminder multiplies the cost of that send.
Sending to the entire base. Without segmentation, you pay for those who would never respond.
Reopening conversations unnecessarily. Each reopening outside the window is a charged message; concluding the matter immediately is cheaper.
Not cleaning the base. Invalid numbers consume sends and generate nothing.
Ignoring free entry points. Not using conversation ads and official buttons means giving up a no-cost service window.
Planning campaigns without the calculation. Approving sends to 10,000 contacts without calculating cost or return is the fastest path to a bill no one can explain.
Comparing only by message price. The relevant cost is per sale generated. A more expensive, better-segmented campaign usually ends up cheaper overall.
Once you understand the logic, Meta’s bill is usually the smallest line item in WhatsApp operation costs — much smaller than platform subscription fees and infinitely smaller than the cost of losing conversations due to poor organization. What drives up costs is not the message price: it’s sending the wrong message, to the wrong list, in the wrong category.
Frequently asked questions
How does Meta charge for WhatsApp messages?
Charges apply to company-initiated template messages, with different prices per category: marketing is the most expensive, utility is much cheaper, and authentication has its own price. Customer-initiated conversations and replies within the service window do not generate message costs. Prices vary by country and are updated periodically, so consult the current official table for Brazil before planning campaigns.
Is there a cost to reply to customers on WhatsApp?
When the customer initiates the conversation, service within the window does not generate message costs. This means operations that get the customer to start the conversation — via website button, ad link, QR code, or Google profile — spend much less than those based on sending messages proactively. Conversations originating from ads that open WhatsApp often have an extended free window.
How to reduce WhatsApp message costs?
Five measures solve most of it: correctly classify templates (don’t insert offers in transactional messages, which would turn them into marketing), encourage customers to initiate conversations, resolve issues within the open window to avoid reopenings, segment campaigns instead of sending to the entire base, and keep the base clean of invalid, duplicate, and opted-out numbers.
Centralize service without changing your number
QR or official API connection, single queue with assigned owner, client history stored by the company, and sales funnel on the same WhatsApp dashboard.