How Much Does a WhatsApp Bot Cost? 2026 Prices ($1K-$3.5K)

Meta publishes the current categories, rates, and effective dates on its official WhatsApp pricing page. Use that page for each quote because the billing model changes over time.

“Make is a no-code platform for connecting apps and automating workflows visually.” — Make.com official site

The cost of a WhatsApp bot splits into three parts. Most people only think about the first one, then get caught off guard by the other two — so here’s each one in plain terms.

Part 1: The one-time setup. This is what you pay once to have the bot built and switched on. It covers mapping out every situation the bot should handle, getting it connected to WhatsApp the official way through the WhatsApp Business API (which includes getting your business approved by an authorized WhatsApp provider, known as a BSP, if you go that route), writing out how each conversation flows, hooking it up to the system where you keep your customers, testing everything, and helping you go live. This usually runs $1,000-$3,500, depending on how much the bot needs to do.

Part 2: The monthly running cost. This can include the WhatsApp connection, workflow infrastructure, model usage, licences and managed operation. Each provider uses different units and can change its plans. Compare the current official rate pages with the proposal’s volume assumptions. Achiya’s current managed operating fee is ₪100-300/month; provider charges are separate.

Part 3: Provider and message fees. The official API uses market- and category-specific pricing that changes over time. A business solution provider may add its own markup or subscription. The quote should state the rate-card date, expected message mix, and provider fee separately.

What you get: answers for a defined FAQ set, opening-hours logic, a menu or service list, and lead collection. It can use WAHA or the official API, depending on policy and reliability requirements. Delivery is defined after the scenarios and connection are mapped. See the clinic guide and restaurant guide for measurement plans in those industries.

Platform Comparison: WAHA vs Meta Official API vs Twilio vs Wati (2026)

The table below lays out the four ways we most often connect a bot to WhatsApp for small and medium businesses. We checked every price against the providers’ own pages in May 2026.

There is no universal volume tipping point. Price the actual message mix, provider fees, hosting, maintenance, and migration risk. An unofficial connection can be restricted regardless of volume.

For a deeper dive into the BSP landscape, see our WhatsApp Business API guide and the WAHA vs official API comparison.

WhatsApp Bot on a Monthly Subscription (SaaS)

Alongside custom builds, Israeli platforms sell bots as recurring subscriptions. Ask each provider for a current written quote with the same scenarios, integrations, message volume, data export, support, and usage limits. Compare that with the custom-build setup and operating costs over the same period. The contract, not the payment model alone, determines what can be exported or moved later.

When is a subscription still the better call? A one-or-two-month trial, a temporary campaign, or genuinely zero setup budget. In those cases check two things up front: the exit cost (can you export your conversations and bot logic?), and what happens to the price once you cross the plan’s conversation cap.

Choice 1: WAHA or the official API. WAHA has no software license fee but needs hosting, maintenance, and an explicit decision about unofficial-provider risk. The official API has current message and provider fees. Compare them with the project’s actual volume and support requirements.

Choice 2: the tool that runs the automation. This is the engine working behind the scenes to make the bot actually do things. One option, n8n, is free and open to everyone — you just run it yourself on a rented cloud computer, so it costs $0. The other, Make.com, is a ready-made service: its free plan covers up to 1,000 actions a month, which is enough for a quiet bot, and its next plan up is $9/month for 10,000 actions. Zapier does much the same job but usually costs 2-3 times more than Make.com. Our rule of thumb for small businesses: pick n8n if you’re comfortable with tech, and Make.com if you’d rather not deal with it.

Choice 3: managed operation or project work. The proposal should state what the monthly operating fee covers and what counts as a separately quoted change. Under Achiya’s current pricing, managed operation is ₪100-300/month; new features and third-party charges are separate.

1. How many situations it handles — each new type of conversation the bot needs to manage adds $100-300 to the build. A bot that deals with 5 situations really is cheaper to make than one that deals with 20.
2. What it connects to — tying it into your customer list adds $300-800, depending on how tricky that system is. Letting it take payments adds $500-1,000. Booking into your calendar is usually already part of the business package.
3. How smart it is — a simple bot that just reacts to keywords is included in the basic tier. Giving it real AI that understands plain language (the technology behind tools like ChatGPT, called a large language model, or LLM) adds $1,000-2,000.
4. How many messages a month — sending more costs more to run, but it can also earn you a better price per message if you agree on a volume deal with your WhatsApp provider.
5. Your own look and voice — getting your branded messages approved, writing custom greetings, and keeping the same personality in every reply — usually $200-500.
6. How many languages — each extra language adds $300-600, since every conversation has to be translated and each message re-approved in that language.
7. Reliability and monitoring scope — define the availability target, monitoring hours, alert path, response responsibilities, and provider dependencies in writing.

The biggest hidden factor of all: how clearly you know what you want before work starts. Someone who shows up with a clear one-page list of everything the bot should do tends to pay noticeably less than someone who begins with “we’ll figure it out as we go.” When it’s spelled out from the start, the work goes faster and nothing has to be rebuilt halfway through. When it’s fuzzy, the job drags on and usually costs the same or more.

In plain words, the little formula below adds up what the bot saves you and earns you each month, subtracts what it costs to run, and then shows how many months it takes to earn back the setup fee — if math isn’t your thing, the three steps right after it walk you through each piece:

Net monthly benefit =
  (monthly hours saved × hourly cost)
  + (additional revenue from 24/7 availability)
  − (monthly operating cost)

Payback in months = Setup cost ÷ Net monthly benefit

Step 1: measure handling time. Record how many minutes the team spends on the selected WhatsApp tasks before launch. After launch, use the same sampling method and include exception handling and review time.

Step 2: measure attributable outcomes. Define a completed booking, qualified lead, or order and keep the attribution rule fixed. Do not assign revenue to an after-hours message unless the data connects it to a completed outcome.

Step 3 — Subtract the actual operating cost. Use the ₪100-300/month managed operating fee from the proposal, then add the current message, model, licence and other third-party charges that apply to the business’s volume.

Treat the calculated payback as a measured result, not a guarantee. If net monthly benefit is zero or negative, there is no payback period to report. Use Meta’s official pricing for message costs and document every assumption.

Mistake 2: Forgetting to count the per-message cost. If you’re firing off 10,000 or more promotional messages a month, those tiny per-message charges add up to $300-500/month on their own — sometimes more than the whole setup fee, every three months. Figure out roughly how many messages you’ll send before you decide which way to connect.

Mistake 3: Not writing down what you want. Vague scope makes price and delivery harder to control. Bring a one-page list of scenarios, integrations, expected volume, and acceptance criteria.

Mistake 4: Not asking who fixes it later. Some quotes throw in 30 days of help after launch; others include none at all. A bot that works great on day one but falls over the moment Meta changes its rules next month isn’t worth as much as it looks. Always ask what happens afterward.

Mistake 5: Not checking whether it actually pays off. Don’t get a bot just because “everyone has one.” Get it because the numbers make sense for how many messages you get. If they don’t, answering by hand with the free WhatsApp Business App is a perfectly good place to start.

• Meta message rates and billing rules changed. Check the current market and category on Meta’s official pricing page before using an old quote.
• WhatsApp phone calls are now open to businesses (March 2026) — with the official API, your system can now place and answer calls over WhatsApp on its own. This isn’t possible with WAHA or the unofficial options. So if you want the bot to make calls at set times, like confirming appointments or following up, that pushes you toward the official API.
• Human handoff scope differs by platform. Compare limits and billing on the providers’ current pages; our Chatwoot vs Intercom comparison explains the fields to check. (Reader perk: 5% off Chatwoot Cloud with code ACHIYAVS.)
• AI model costs differ by model and workload. Use the current Anthropic pricing and other provider pages with your expected token volume.