WhatsApp chatbot pricing in India: platform fees vs Meta's per-message charges

Updated 2026-09-19 · 5 min read · By the Convo360 team

Two bills, not one

Every WhatsApp chatbot bill in India has two parts, and confusing them is the most common pricing mistake businesses make. First, the platform fee: what you pay a provider like Convo360 for the flow builder, AI agent, CRM, inbox and automation. Second, Meta's per-message charges: billed directly by WhatsApp for each message delivered, at the rate for its category, completely independent of which platform you use.

On Convo360's BYOK model, your WhatsApp number is your own, so Meta bills you directly at their published rates, with no markup layered on top by the platform. Read what is the WhatsApp Business API if you're new to the underlying system.

Meta's message categories explained

Every template you send is a marketing, utility or authentication message, free-form replies inside the window are service messages, and Meta prices each per delivered message (conversation-based pricing was replaced by per-message pricing on 1 July 2025). Read the details in our template categories guide and 24-hour window explainer, and check Meta's own WhatsApp Business Platform pricing page for current published rates.

What replies inside the 24-hour window cost

When a customer messages you first, a 24-hour customer service window opens. Any free-form reply you send inside that window (no template required) is a service message, and that includes replies written by an AI agent. Service messages are free until 30 September 2026. From 1 October 2026 Meta charges them per message at the market's utility rate (INR 0.115 for India on Meta's October rate card) after a free tier of 1,000 delivered service messages per business phone number per month, with no roll-over. Utility templates sent inside an open window are also charged from that date, and without a payment method in Meta Billing Hub, service messages beyond the free tier are not delivered. Full details are in WhatsApp Business API pricing in India.

The moment that window closes and you need to reach the customer again, you must send an approved template, which is charged per delivered message at the rate for its category.

Worked examples: 1,000 and 10,000 conversations a month

Take a salon replying to roughly 1,000 customer-initiated conversations a month: mostly booking and rebooking questions answered comfortably within the active window, plus a small number of utility reminder templates for appointments. Until 30 September 2026 those replies are free, so Meta's charge is driven by the reminder templates. From 1 October 2026 each reply is a service message: the first 1,000 delivered per number each month are free and the rest are charged at the utility rate, so count messages, not conversations. The software side is simple here: the Starter plan (₹999/month, ex-GST) covers WhatsApp, the flow builder and the knowledge base at this volume.

Now take a D2C brand running festive broadcasts to 10,000 contacts, mixing marketing templates for the campaign with utility templates for order updates. That sits comfortably on the Pro plan (₹2,999/month, ex-GST) for the software layer, which also unlocks AI voice calls and all integrations. At this volume, Meta's per-message charges become the larger line item, since every marketing template delivered is charged, even to a recipient who is inside an open service window. Plan your broadcast campaigns around this: a well-segmented list of 10,000 engaged contacts often costs less overall than an unsegmented blast to 20,000, even though the platform fee doesn't change. Check Meta's current published rates for the exact per-template figure, since rates vary by category and are set by Meta, not the platform.

How to keep the Meta bill down

• Reply within the 24-hour window using free-form messages instead of templates wherever the customer messaged first.
• Use Utility templates instead of Marketing for anything transactional, such as order status or appointment reminders, since utility messages are priced lower than marketing messages.
• Segment broadcast lists so you're not messaging contacts unlikely to convert, since each delivered template is billed.
• Check Analytics monthly to see which flows are driving the most template sends, and tighten them.
• Avoid re-sending a marketing template to someone who already replied: a follow-up free-form message inside the window is a service message, free until 30 September 2026 and charged at the lower utility rate, after the monthly free tier, from 1 October 2026.

Choosing the right Convo360 plan

Yearly billing saves roughly 10% across all plans. See Billing docs for GST invoicing details and how GST invoices work for SaaS subscriptions.

Mistakes to avoid, and a checklist before you sign up

Common mistakes when estimating this cost:

• Assuming the platform fee is the whole bill and being surprised by Meta's invoice at scale.
• Using Marketing templates for transactional updates that should be cheaper Utility templates.
• Not tracking which flows keep sending paid templates unnecessarily.
• Forgetting that broadcast volume, not chatbot complexity, is usually what drives the Meta bill up.
• Comparing platforms only on the software fee, while ignoring whether a competitor is quietly marking up Meta's own per-message charge.

Before you sign up, work through this checklist:

• Estimate whether your volume is mostly customer-initiated (support, replies) or business-initiated (reminders, broadcasts): this decides which category dominates your Meta bill.
• Pick Utility over Marketing wherever the message is transactional, not promotional.
• Check that the platform you choose is genuinely BYOK, so Meta's rate reaches you unmarked-up; see per-message markup vs flat fee for how to spot the difference.
• Match the Convo360 plan to the features you need (AI voice, integrations, API) rather than to message volume, since volume is billed separately by Meta.

Example conversation

Frequently asked questions

Related reading

Available across India — see WhatsApp & AI calling automation by state and city.