- Key, token and route: the three layers of limit
- A quick sum before you build a sync
- Reading the 429 you got back
- Members and search routes have tighter rules
- Using the response headers to stay under
- Polling less: webhooks and batching
- How Asana, Jira and Pipedrive count requests
- My advice for anyone integrating with Trello
- Frequently asked questions
Trello's API allows 300 requests per 10 seconds for each API key and 100 requests per 10 seconds for each token. Requests to /1/members are also limited to 100 requests per 900 seconds. Go over and you'll get a 429 error, so use webhooks and nested resources instead of heavy polling.
Key, token and route: the three layers of limit
Trello counts your requests in three overlapping ways, and the one that bites first is rarely the one people expect.
| Item | Limit | Applies to |
|---|---|---|
| Requests per API key | 300 per 10 seconds | Each API key |
| Requests per token | 100 per 10 seconds | Each token |
| Requests to /1/members/ | 100 per 900 seconds | Special route limit |
| 429 errors per key | Over 200 triggers 429 for the rest of the 10-second window | Each API key |
| Webhooks | No limit on the number you can set up | All |
The token limit is the tighter one. A single token can make only 100 requests per 10 seconds, even though the key allows 300. So an integration that runs everything through one user's token hits the wall at a third of what the key suggests.
A quick sum before you build a sync
Say your integration checks 20 boards every minute and makes five calls per board. That's 100 calls in one burst. On a single token, that's the entire 10-second allowance gone in one go, and any other call that user's token makes in the same window gets a 429. Spread those calls across the minute, or switch the checks to webhooks, and the same job fits comfortably. Do this sum for your busiest moment, not your average one.
Reading the 429 you got back
Every request over a limit returns a 429 with a message saying which limit was exceeded. That's more helpful than many APIs, so log the message rather than just the status code. Here's what each one means.
- API_TOKEN_LIMIT_EXCEEDED: the token has gone past 100 requests in 10 seconds. Slow that user's calls or spread the work.
- API_KEY_LIMIT_EXCEEDED: the whole key has gone past 300 in 10 seconds, usually because several tokens are busy at once.
- API_TOKEN_DB_LIMIT_EXCEEDED: the token is making too many calls that are heavy on Trello's database. Back off the large requests.
- API_TOO_MANY_CARDS_REQUESTED: you asked for all of a board's cards with all their actions in one go. Fetch the cards first, then their actions.
Watch out for retry storms. If the 429 count for a key goes over 200, Trello returns 429 for every request for the rest of that 10-second window, so a script that retries instantly can lock itself out.
Members and search routes have tighter rules
The Atlassian Trello developer page on rate limits flags three routes with special, stricter limits: /1/members, /1/membersSearch and /1/search. A call to /1/members/ counts against the members limit, the key limit and the token limit all at once. If your app looks up members on every run, that 100 per 900 seconds will run out long before anything else. Cache member details and refresh them occasionally instead.
Using the response headers to stay under
Trello returns rate limit headers such as x-rate-limit-api-token-max and x-rate-limit-api-key-remaining. Read them on every response and slow down as the remaining count drops, rather than waiting for a 429 to tell you. It's the simplest way to make an integration polite.
Polling less: webhooks and batching
The support guidance is clear. Use webhooks rather than polling, and there's no limit to the number of webhooks you can set up. Where you must poll, lower the frequency and batch calls about every minute. Use nested resources to fetch related data in one call instead of several.
If you're not sure when a webhook beats polling, my explainer on APIs versus webhooks walks through it, and if you're building for a marketing team, my Trello marketing guide shows what's worth automating.
How Asana, Jira and Pipedrive count requests
Asana limits its search API to 60 requests per minute, on top of its general limits.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Jira Cloud uses a points-based model where each call uses points based on the work it performs, and returns 429 when a limit is exceeded. The details are in Jira's points, burst and per-issue caps.
Pipedrive uses a daily token allowance plus burst limits over a rolling 2-second window.
My advice for anyone integrating with Trello
Switch to webhooks for anything that watches for changes, cache member lookups, and read the remaining-requests header on every call. If one token does all the work, it'll hit the 100-request limit first, so spread work across tokens where that makes sense. And never retry a 429 straight away.
Want your tool on this list? Featured spots and reviews here are paid partnerships. Here's how to partner with me.
Frequently asked questions
What's the Trello API rate limit?
300 requests per 10 seconds per API key and 100 requests per 10 seconds per token.
What happens if I exceed it?
Trello returns a 429 error naming the limit exceeded.
How do I avoid hitting it?
Use webhooks instead of polling, and nested resources to fetch related data in one call.
Is there a limit on webhooks?
No. Trello says there's no limit to the number of webhooks.
Where can I see how many requests are left?
In the x-rate-limit response headers that come back with each call.