Skip to content

DM Rate Limits and Throttling

Updated

Meta decides how fast your account may send messages, not LinktoDM. Push past that line and the platform stops accepting messages — and an account that keeps pushing looks like spam, which is the more expensive problem of the two.

The rate shown in the queue panel therefore isn’t a fixed setting. It’s the speed LinktoDM has judged safe for your account at this moment.

The line under the queue panel — System will automatically reduce delivery speed during peak times — is the feature announcing itself, and Learn more beside it spells out the reasoning: when Instagram’s rate limits are hit, slowing messages down improves delivery success and avoids tripping spam detection.

In short, LinktoDM watches how the platform responds and eases off when it sees rate limiting, then picks the pace back up. Sending speed is adjusted for you, in both directions, and there is nothing to switch on.

The limits aren’t a secret — Meta publishes them in its Graph API rate limiting documentation. They’re worth knowing, because they explain why the queue behaves the way it does. The figures below are Instagram’s, which is where most LinktoDM automations run; Facebook Page messaging sits under the same system with its own numbers, and the queue treats both the same way. Instagram applies them per professional account, so every account you connect has its own budget:

What it covers Instagram’s limit
Your account’s total API activity 4,800 calls per 24 hours × the impressions your content earned in that window
DMs sent as a private reply to a comment on a post or reel 750 per hour
Private replies to comments on a Live 100 per second
DMs with text, links, reactions or stickers 100 per second
DMs carrying audio or video 10 per second

Two of those numbers do most of the work in practice.

The 24-hour budget moves with your reach. It’s a rolling window, and it’s multiplied by impressions — so a post that takes off raises your ceiling at the same time as it raises your volume, while a small account working through a backtracked week of comments has much less room than its follower count suggests.

Comment-to-DM automations meet the 750/hour wall first. The DM that answers a comment on a post or a reel is a private reply, and private replies are capped by the hour rather than by the second. A viral post generating thousands of qualifying comments in an hour is over that line no matter how the sending is spaced — which is exactly the case the queue is built for: it holds the surplus and keeps releasing it into the next hour instead of throwing it away.

When an account does cross a limit, Instagram answers with a throttling error (code 80002 for Instagram rate limits, or 4, 17, 32 and 613 depending on the token and the call) and reports how long access is restricted for. LinktoDM reads those responses and slows the queue rather than retrying into a closed door.

Slower is the safe outcome, not the failure. When the platform starts rate limiting, pushing harder gets messages rejected; easing off gets them delivered. A queue that drains slowly during a spike is the system choosing delivery over speed — the messages still arrive, a little later than they would have.

There’s nothing to configure. Throttling is automatic. There’s no setting to raise the rate, no way to buy more of it, and no priority lane for one message. The published limits are Instagram’s, and they apply to every tool that sends on your behalf.

Peak traffic is when it shows. A reel doing numbers, a launch, a giveaway — the hours when your automations are busiest are the hours the platform is strictest.

If the queue backs up further than you’re comfortable with, the lever is upstream of the queue:

  • Tighten the trigger so fewer comments qualify in the first place.
  • Run fewer automations on the same account at once, especially on posts that are live at the same time.
  • Cap the public replies with a per-post reply limit, which keeps a single post from generating an unbounded run of activity.
  • Be deliberate with backtracking: it answers a week of comments in one go, which lands in the queue as one burst — against an hourly private-reply limit that doesn’t care how old the comment is.