WhatsApp BSP Fees: What Providers Add to Meta's Rates
You have two WhatsApp API quotes on your desk. One is a monthly plan with "Meta fees at cost". The other is a single price per message with no subscription. They look impossible to compare, and that is partly by design.
Every provider sits on the same base: Meta's published per-message rates. WhatsApp BSP fees are everything a Business Solution Provider adds on top, and providers package them very differently: subscriptions, markups, per-number licences, seats, setup fees and wallet rules. Once you separate Meta's part from the provider's part, comparing quotes becomes arithmetic.
This guide covers the common fee models, the smaller line items that change the total, a worked comparison, and the questions to ask before you sign.
Two layers: Meta's rate and the provider's fee
Meta charges per delivered message, at a rate set by the message category and the recipient's country. That rate card is public, and the list rate is the same no matter which provider you use. Our WhatsApp API pricing guide covers it in detail.
The provider's fee is the second layer, and it pays for what the provider adds: onboarding, an inbox, template tools, chatbot builders, analytics, support and often credit terms.
The money can flow two ways. Either the provider pays Meta and re-bills you, or your messaging account carries your own payment method and Meta bills you directly while the provider charges separately for its software. Ask which applies. It decides whose currency you pay in, who sees Meta's invoice, and how easily you can leave.
The five common fee models
Most quotes are one of these, or a mix:
| Model | What the invoice shows | Watch out for |
|---|---|---|
| Platform subscription, Meta at cost | Monthly fee plus Meta's rates passed through | Feature gating between plans; API access only on higher tiers |
| Per-message fee on top | Meta's rate plus a fixed fee per message | Fees on inbound messages and on failures |
| Per-number licence | Monthly fee per WhatsApp number plus Meta's rates | Costs scale with numbers, not usage |
| Blended per-message rate | One price per message, Meta's rate hidden inside | The markup is invisible and can differ by category |
| Seats on top | Per-agent price for the inbox or CRM, plus one of the above | Seat minimums and annual commitments |
Two public examples show how different the shapes can be. Twilio lists a fee of $0.005 per message, inbound or outbound, on top of Meta's fees, which it passes through, plus $0.001 for each message that ends in a failed status, per its WhatsApp pricing page as of October 2026. 360dialog prices its WhatsApp API plans per number per month, starting at €49, plus Meta's messaging fees, per its pricing page as of October 2026. One scales with messages in both directions; the other scales with numbers. Neither is cheaper in general. It depends on your volume and how many numbers you run.
The smaller line items that change the total
Headline fees rarely tell the whole story. Look for:
- Setup or onboarding fees, sometimes framed as help with business verification or a green tick application.
- Inbound message fees. Meta never charges for messages customers send you; some providers do.
- Failed message fees. Meta charges only for delivered messages; some providers charge for failures.
- Prepaid wallets with minimum top-ups, expiry dates or non-refundable balances.
- Currency conversion. If the provider bills in a different currency from its Meta account, a conversion margin can sit inside the rate.
- Add-ons: chatbot sessions, Flows or form builders, extra numbers, extra seats, API access or webhooks on a higher plan only.
- Commitments: annual contracts, minimum monthly spend, or notice periods.
- Taxes, which may or may not be included in a per-message quote.
Then there are the October 2026 changes. Service replies are now charged after 1,000 free per number per month, and utility templates sent inside the customer service window are charged too. A provider can pass the free allowance through cleanly, mark up the new charges, or bundle them into a blended rate. Ask how yours does it.
Reading a quote: a worked comparison
To compare quotes, rebuild each one as Meta's cost plus the provider's cost at your real volumes. The numbers below are examples for illustration only, not any provider's or Meta's prices.
Suppose your market's Meta rates were $0.05 per marketing message and $0.005 per utility or service message, and each month you send 20,000 marketing templates, 10,000 utility templates and 6,000 service replies on one number. Neither example provider charges for inbound messages.
| Provider A | Provider B | |
|---|---|---|
| Pricing model | $99/month subscription, Meta at cost | Blended: $0.065 marketing, $0.007 utility and service |
| Meta's share | $1,000 + $50 + $25 = $1,075 | Same $1,075, inside the blended rates |
| Provider's share | $99 | $300 + $20 + $10 = $330 |
| Monthly total | $1,174 | $1,405 |
In Provider B's case, the provider's share is the blended rate minus Meta's rate, multiplied by volume: $0.015 on each marketing message and $0.002 on each utility or service message (only 5,000 service replies are billable after the free 1,000). The gap grows with marketing volume: every extra 1,000 marketing messages adds $15 to Provider B's share and nothing to Provider A's.
Run the same exercise with your own volumes using our WhatsApp API cost calculator, and always confirm Meta's rate for your markets on the official rate card before you subtract it.
Questions to ask every provider
- Show me Meta's rate and your fee as separate lines, per category and market.
- Do you charge for inbound messages, failed messages or both?
- How do you bill service messages since October 1, 2026, and do I get the 1,000 free per number?
- Do volume tier discounts on utility and authentication reach my invoice?
- Is billing through your account or my own payment method on the messaging account? Which currency?
- For India: has my messaging account been moved to INR ahead of Meta's December 31, 2026 deadline?
- Who owns the business portfolio and the messaging account? Can I migrate my number to another provider, and what does that cost?
- Which features need a higher plan: API access, webhooks, extra numbers, chatbots?
- Is there a setup fee, a minimum monthly spend or an annual commitment?
- How do wallet balances work if I leave?
On migration: Meta supports moving numbers between providers, and eligible templates are duplicated into the destination messaging account, so a provider shouldn't hold your templates hostage. Our guide to switching WhatsApp API providers walks through the process.
When going direct makes sense
You can skip the provider layer by connecting to Meta's Cloud API yourself. You then pay only Meta's rates, but you build and run everything a provider would supply: webhook hosting, template management, an inbox for your team, monitoring and support escalation. That trade usually pays off for companies with engineers to spare and a narrow use case, such as OTPs or order notifications from an existing system. Our comparison of Cloud API vs a BSP goes deeper, and our roundup of the best WhatsApp API providers compares named options. For the whole platform, see the complete guide to the WhatsApp API.
Where Tenashi fits
Many teams shopping for a BSP don't actually need templates or the green tick. They need their team to answer WhatsApp together, track deals and stop losing chats on one person's phone. For that, Tenashi is a different kind of product with a simpler bill. It is not a BSP and doesn't sell Meta API access. You link the WhatsApp or WhatsApp Business number you already use with a QR scan, and your team gets one shared inbox for DMs and groups, a CRM under every chat, tickets and follow-ups, at a flat monthly price from $18/mo with no per-message fees, template approvals or wallets.
The trade-off is real. Without the official platform there are no templates, interactive buttons or OTP-grade throughput, and sends are paced to protect your number. WhatsApp can still ban numbers that spam; ban protection reduces that risk without guaranteeing anything.
See every feature · Get started
FAQ
What is a WhatsApp BSP fee?
It is anything a Business Solution Provider charges on top of Meta's per-message rates: a monthly subscription, a markup or handling fee per message, a per-number licence, per-agent seats, setup fees or add-ons. Meta's rates are public and the same everywhere, so the BSP fee is the part of your bill that actually differs between providers.
Do BSPs mark up Meta's WhatsApp rates?
Some do and some don't. Many pass Meta's rates through at cost and charge a subscription or licence instead. Others quote a single blended price per message that includes a margin. Ask for Meta's rate and the provider's fee as separate lines, then subtract Meta's published rate to see the real markup.
Is it cheaper to use the WhatsApp Cloud API directly?
On fees, yes: you pay only Meta's rates. On total cost, it depends. Going direct means building and maintaining webhooks, template management, a team inbox and monitoring yourself. For a narrow integration with engineers available, direct is often cheaper. For a team that needs a ready inbox and support, a provider can be.
Do BSPs charge for incoming WhatsApp messages?
Meta never charges for messages customers send you, but some providers charge their own fee on inbound messages as well as outbound. Twilio, for example, lists its per-message fee as applying inbound or outbound. Check every quote for inbound and failed-message fees, because they scale with customer activity rather than with what you send.
Can I switch BSPs without losing my number?
Generally yes. Meta supports migrating a phone number and messaging account between providers, and eligible templates are duplicated into the destination account. Confirm before signing who owns the business portfolio and messaging account, and whether the provider charges for migration, since contracts and ownership decide how smooth a move is.
Separate Meta's rate from the provider's fee and every quote becomes comparable. If you need a shared inbox rather than a template pipeline, see how Tenashi works.