🔧

Make error

Fix “Connection timed out / operation failed” in Make

A Make connection timeout means a module waited for the target service and got nothing back within Make's time limit — the target API is slow or down, a firewall or IP block is refusing Make's servers, the connection's authorization has gone stale, or the request is too large to finish in time. Check the target service first, re-authorize the connection, then shrink or split the request.

Timeouts in Make feel vague because the error only tells you the conversation ended badly — not which side hung up. But there are only a few possibilities: the target service was slow or down, Make's servers couldn't reach it at all, the saved authorization had expired, or the request was simply too big to complete within the limit. This page shows how to tell those apart quickly and what to change for each, so the fix sticks instead of the timeout returning next week.

What causes it — in the order they happen

The target service is slow or having an outage

You'll know it's this if: The module worked yesterday, fails now, and other scenarios using the same service fail too.

The fix: Open the target app in a browser and confirm it's working normally, then check its status page. If the service is degraded, wait and re-run — Make can retry failed modules, and a re-run after the service recovers usually succeeds. Don't rebuild the scenario while the outage is still on.

Make's servers can't reach the target

You'll know it's this if: The timeout happens on every run, including with the tiniest request, while the app works fine in your browser.

The fix: Make runs from its own cloud servers, not your computer — firewalls, IP allowlists and geo-blocks that permit your machine can still refuse Make. If the target service has IP allowlisting or bot protection, add Make's official IP ranges (documented at help.make.com) or an approved integration method. Test with a minimal request to confirm.

The connection's authorization has gone stale

You'll know it's this if: The module hangs or fails after working for weeks, sometimes right after a password change or an admin revoked access.

The fix: Open the module's connection settings and re-authorize the account. OAuth tokens expire or get revoked silently, and a connection that looks present can no longer complete a handshake. Reconnecting takes a minute and resolves a large share of suddenly-timeoutting modules.

The request is too large to finish in time

You'll know it's this if: The timeout appears only on runs with lots of data, while small test runs succeed.

The fix: Shrink the request: reduce the limit on a search module, use pagination so each operation handles one page, and filter early so only the records you need travel through the scenario. Modules have a fixed time budget per operation — a request that needs minutes will never fit, no matter how you configure it.

The target API is rate limiting or queueing you

You'll know it's this if: Timeouts cluster when the scenario runs many operations in a burst, or after another system started hitting the same API.

The fix: Slow the scenario down: lower the maximum number of cycles, add a sleep between operations, or split the work across scheduled runs instead of one big batch. If the service documents rate limits, keep the scenario comfortably under them rather than testing the edge every run.

The wrong URL, port or endpoint

You'll know it's this if: The timeout is immediate and every run, on a self-built HTTP module or a recently edited one.

The fix: Check the exact URL in the module: a typo'd domain, a missing https://, a closed port, or an endpoint that's slow by design will hang until Make gives up. Test the same URL from a tool like a browser or curl to confirm it responds, then correct the module's URL.

Fix it step by step

1
Confirm the target service is actually up
Use the app normally in your browser and check its status page. If it's down or degraded, schedule a re-run and move on — no configuration change fixes a service that isn't answering.
2
Re-authorize the connection
Open the failing module, go to its connection, and reconnect the account. This clears expired or revoked tokens, which is the most common cause of a module that worked for weeks and suddenly times out.
3
Test with the smallest possible request
Run the module on one record with the fewest fields. If a minimal request succeeds but the real run times out, the problem is size or rate — move to the next step. If even the minimal request times out, the problem is reachability — check IP allowlists and the module's URL.
4
Shrink, paginate or split the work
Reduce search limits, add early filters, and handle large datasets in pages rather than one request. If one scenario does hours of work, split it into scheduled chunks so each operation stays comfortably inside its time budget.
5
Add an error handler for the stragglers
Attach an error handler route to the fragile module with a Break/Resume or retry setup, so a single timeout doesn't kill the whole scenario — it pauses, waits and continues. Check the scenario history afterwards to see how often it's actually needed.
6
Re-run from history and watch the next scheduled run
Re-run the failed operations from the scenario history once the fix is in, and confirm they complete. Then check the next scheduled run — a timeout that's truly fixed doesn't come back on real data.

Stop it coming back

Re-authorize connections whenever a module starts failing after a long period of working — stale tokens are the most common sudden timeout.
Design big jobs as paginated, scheduled chunks instead of one massive batch request.
Keep scenarios under the target API's documented rate limits; add sleeps between bursts of operations.
Add a retry or break/resume error handler to modules that talk to slower or less reliable services.
If your target service has IP allowlisting, add Make's published IP ranges before the first run — don't discover the block in production.

Still stuck on the exact screen?

Paste a screenshot of the error into AI Online Office and it tells you what to click next — on your screen, in seconds.

Fix it with AI Online Office

Make error FAQ

What does a connection timeout mean in Make?

It means a module sent a request to another service and got no response within Make's time limit. The service was too slow or down, Make's servers couldn't reach it, the saved authorization had expired, or the request was too large to finish in time.

Why does my Make scenario time out when the app works fine in my browser?

Make runs from its own cloud servers, not your computer. Firewalls, IP allowlists and bot protection that let your machine through can still block Make's servers. Check the service's allowlist against Make's published IP ranges.

How do I increase the timeout in Make?

Most modules have a fixed per-operation time budget you can't extend. The practical fix is to make the request smaller: paginate, reduce limits, filter early, and split large jobs into scheduled chunks. For long-running external processes, use a webhook-based pattern where the service calls you back when it's done.

Why did my Make connection stop working suddenly?

The most common cause is an expired or revoked OAuth token — the connection still looks set up but can no longer authenticate. Re-authorize the account in the module's connection settings; it takes about a minute.

Should I just add a retry to every module?

No — add error handlers to the modules that genuinely talk to slower or rate-limited services. A blanket retry hides the difference between "the service hiccuped" and "my request is malformed", and the second one will never succeed no matter how many times you retry.

Other errors we fix

New to the tool rather than the error? Start with the plain-English guides, or read what an AI Online Office is.