What is the WhatsApp pair rate limit?

How the pair rate limit works

Every send on the Cloud API is a pair: one business number sending to one customer number. Meta's pair rate limit looks only at that specific pair, roughly one message every 6 seconds, and is separate from the number-wide messaging limit, which counts unique customers reached outside the 24-hour window across the whole business portfolio. A number can be well inside its messaging limit for the day and still hit the pair rate limit if it sends several messages to the same customer within a few seconds.

This is different from Meta's general per-app and per-account request-rate limits, which cap how many API calls your integration can make overall. The pair rate limit is about the pace of messages to one recipient specifically, whatever your total call volume looks like.

Why integrations hit it

• Retry loops. A webhook handler that resends a message immediately after a timeout, without checking whether the first send actually failed, can fire two or three sends to the same customer inside a few seconds.
• Chatbot steps that fire too close together. A flow with several Send Message steps in a row and no delay between them can breach the limit for a fast-moving conversation.
• Duplicate automations. Two systems (a CRM trigger and a chatbot step, say) both reacting to the same event and messaging the same customer at the same moment.

Fixing it

1. Add a short delay (a few seconds is usually enough) between consecutive sends to the same customer, rather than firing them back to back.
2. Check the result of a send before resending; a timeout on your side does not always mean the message failed to send on Meta's side.
3. Make sure only one system is responsible for messaging a given customer at a given step, so two automations do not race each other.
4. On error 131056, back off and retry after a short pause rather than retrying immediately, which only repeats the problem.

Pacing in Broadchat

Broadchat's own campaign pacing, under CRM → Bulk Campaign, already spaces sends to stay within a number's messaging limit and throughput, and retries automatically when the pair rate limit is hit during a bulk send. The limit is more often seen by developers calling the permanent access token or a chatbot's API step directly; if your own code sends outside Broadchat's campaign engine, build the delay and retry logic described above into that code.