New Airtable API call limits

New Airtable API Call Limits

Airtable has announced that their FREE plan will soon be limited to 1,000 API calls per month. (There is no way of seeing in the system how many API calls are made for an account :sleepy:) If the monthly limit is exceeded, they “may” start to limit the number of calls processed per second.

The processing speed of Airtable is not critical to me or my customers. However, I am concerned about how the throttling back at Airtable’s end may affect Make scenarios (mainly Search). Will it cause them to time out, or throw errors?

Usually when a rate limit is reached, the server will respond with a HTTP error code (4xx). A common error code is “429: Too Many Requests”. This is most likely what would happen, and you’ll just need to add an error handler to your Airtable modules to perform an alternative path (or notify you).


I was hoping not to hear that :sleepy:

I have several hundred scenarios and many hundred Airtable modules with them. This means putting in a Retry error path for every module. There is no alternative path possible…It’s either processed or it’s failed.

I assume that there’s no way to tell Make to “slow” down or wait longer?

To sleep, retry, and resume, it takes an additional two operations.

In this error handling example, I duplicated my ChatGPT module after the sleep module, and resumed the response.

The alternative is just to upgrade to a paid plan, which is what Airtable hopes free users will now have to do, because they are a business, and businesses have to make money.


Thanks for that example - I may have to deploy that. After all, what’s several days of coding and testing amongst friends?

I totally understand that Airtable has a commercial imperative. However, changing a plan at relatively short notice, from infinite API calls to only 1,000 - is a MASSIVE jump. It will hurt their goodwill/brand. I will no longer recommend them.

What Airtable is saying is that when the requests reach the monthly threshold, they will not start declining requests - they will simply throttle back on the response time to requests. If that’s the case, they wouldn’t be throwing errors back to Make, just taking longer to respond. What I’m curious to know is, will Make accommodate a delay or will it naturally time out (and if so, how long will that take?)

It depends on how long the delay is.

On a free plan, your scenarios can run up to 5 minutes, and on paid plans up to 40 minutes.


1 Like