Facebook Graph API rate limits: errors 4, 17, 32, 613 and 80001
Facebook's Graph API gives a Page or system user token 4,800 Pages API calls per engaged user in a rolling 24 hours. Past a limit it returns 4, Application request limit reached; 17, User request limit reached; 32, Page request limit reached; 613, Calls to this api have exceeded the rate limit; or 80001, There have been too many calls to this Page account. Stop calling, then resume gradually.
easyCoexistence calls each connected Page with the Page's own token and takes in its feed activity, mentions and messages through Meta's webhooks, forwarding them to your endpoint if you want, so nothing has to poll the Page.
The code when the app whose token is used in the request has reached its rate limit.
The code when the user whose token is used in the request has reached their rate limit.
The code when the user or app whose token is used in a Pages API request has reached its rate limit.
The code for a custom rate limit; with subcode 1996, Meta has noticed inconsistent behavior in the app's request volume.
The Business Use Case code for Page calls made with a Page or system user access token.
Pages API calls allowed per rolling 24 hours with a Page or system user token.
What does each rate limit code mean?
Each code names the limit that was reached, and the token in the request decides which one applies. Meta's rate limiting guide explains what each code indicates, and its error references give the message that comes with it. Here they are side by side, with the message as Meta writes it:
| Meta's message | What reached its limit | |
|---|---|---|
| 4 | Application request limit reached | The app whose token made the request |
| 17 | User request limit reached | The user whose token made the request |
| 32 | (#32) Page request limit reached | The user or app token used in a Pages API request |
| 613 | (#613) Calls to this api have exceeded the rate limit. | A custom limit set by the API being called |
| 80001 | (#80001) There have been too many calls to this Page account. Wait a bit and try again. | Calls made on that Page with a Page or system user token |
Which limit applies to a Page?
It depends on the token. Meta says Page rate limits may use either the Platform or the Business Use Case logic, depending on the type of token used. Pages API calls made with a Page or system user token follow the Business Use Case limit for Pages, while calls made with a user or app token are subject to the user's or the app's Platform limits. When both could apply to a request, Meta applies the Business Use Case limit.
How large are the limits?
Meta publishes formulas rather than fixed numbers, so each allowance grows with use:
- Apps: 200 calls per hour for each daily active user, counted across all of the app's calls rather than per user.
- Pages, with a Page or system user token: 4,800 calls per rolling 24 hours for each user who engaged with the Page in 24 hours.
- Messenger Send API: 300 calls per second per Page for text, links, reactions and stickers, and 10 for audio or video.
- Messenger calls also have a Business Use Case type of their own, reported as error 80006.
How do you see a limit coming?
Read the usage headers. Endpoints that receive enough requests from an app include X-App-Usage, whose call_count, total_cputime and total_time are percentages of the app's Platform limits. Responses rate limited under the Business Use Case logic include X-Business-Use-Case-Usage instead, with the same percentages for each business object, the type of limit, such as pages, and estimated_time_to_regain_access.
X-App-Usage: {"call_count": 28, "total_time": 25, "total_cputime": 25}
X-Business-Use-Case-Usage: {
"{business-object-id}": [
{
"type": "pages",
"call_count": 100,
"total_cputime": 34,
"total_time": 41,
"estimated_time_to_regain_access": 12
}
]
}How long do you wait?
Until the usage drops, which the headers show. For a Business Use Case limit such as 80001, estimated_time_to_regain_access gives the minutes until calls are no longer throttled. For Platform limits, the app's call_count covers a rolling one-hour period, so it falls as older calls age out. Meta's guidance is the same for both: stop making calls once a limit is reached, because continuing to call keeps increasing the call count and the time before calls succeed again.
How do you stay under the limits?
Make fewer, lighter calls and spread them out. Meta's best practices are to spread queries evenly to avoid traffic spikes, to use filters to limit the size of responses and avoid calls that request overlapping data, and, when users are being throttled, to make sure the app is not the cause. Webhooks help too: Meta delivers a Page's feed activity and messages as they happen, so an app need not spend calls looking for them.
Common mistakes
- Retrying at once. Meta warns that continuing to make calls increases the call count and the time before calls succeed again.
- Treating every code the same. 4 is the app's limit, 17 a user's and 80001 a Page's, so slowing down the wrong caller changes nothing.
- Reading X-App-Usage for calls made with a Page token. Those count toward the Business Use Case limit and report in X-Business-Use-Case-Usage.
- Polling a quiet Page often. Its allowance scales with the users who engaged with it, so a Page with little engagement has fewer calls to spend.
When Meta throttles a call, the tool returns Meta's code and message to Claude, ChatGPT or your own code, so your AI can wait before trying again, and a post Meta throttles is retried on its own after 5 minutes, 30 minutes and 2 hours. US$ 9 to US$ 0.75 per connection per month, falling as you add connections, with the first 3 days free.
easyCoexistence is a multichannel API and MCP server that connects AI agents such as Claude and ChatGPT to WhatsApp, Instagram, TikTok, Facebook and more. On WhatsApp it uses Coexistence: the business keeps using its number in the WhatsApp Business app while the AI works through Meta's official API.
Frequently asked questions
What does Application request limit reached mean?
It is error 4: the app whose token made the call has reached its rate limit. Stop making calls, then resume gradually as its usage falls.
What is the difference between error 32 and error 80001?
The token. Error 32 is the Platform limit on a user or app token calling the Pages API; 80001 is the per-Page limit on calls made with a Page or system user token.
How long does a Facebook rate limit last?
Meta sets no fixed time. For 80001, the X-Business-Use-Case-Usage header gives the minutes until calls are no longer throttled; an app's usage is measured over a rolling hour.
What does error 613 mean?
A custom rate limit, set by the API being called. With subcode 1996, Meta has noticed inconsistent behavior in the app's request volume, for example after a recent change in how many requests it sends.
Keep reading
Connect your first channel in two minutes.
No platform reviews to wait for. Connect WhatsApp, Instagram, TikTok or another channel, then ask your AI to post, reply or report, and watch it work.
- 3 days free
- Cancel any time
- Official APIs only
- Monitored 24/7
Verified on