Power Platform Community Forum Thread Details

Maximum RequestTimeoutInMilliseconds value for Copilot Studio connectors

Hello,

I am working with Copilot Studio and configuring a custom connector. I have a question regarding the maximum allowed value for RequestTimeoutInMilliseconds.

According to the documentation, the maximum timeout for connectors should be 30,000 milliseconds (30 seconds). However, when I set RequestTimeoutInMilliseconds to 300,000 milliseconds (300 seconds), I get the following error:

Invalid timeout value configured for Node: yeIJLd, RequestTimeoutInMilliseconds: 300000 milliseconds. Setting timeout to default value: 30000.

Surprisingly, when I set the value to 240,000 milliseconds (240 seconds), it works without any error and the request proceeds as expected.

Could you please clarify:

What is the official maximum value for RequestTimeoutInMilliseconds in Copilot Studio connectors?
Is there any documentation or update regarding timeout values higher than 30 seconds?
Why am I able to use 240 seconds but not 300 seconds?

Thank you in advance!

Categories:

General

Power Automate Flows

Hello フレ-11090352-0!

Great question. You’ve run into a classic “documentation vs. real-world behavior” issue with the Copilot Studio connectors’ RequestTimeoutInMilliseconds setting.

According to the official docs, the maximum timeout is 30,000 milliseconds (30 seconds). If you try higher (like 300,000 ms), you receive an error and it defaults to 30 seconds. However, you noticed that 240,000 ms (240 seconds) works fine, which clearly contradicts the stated limit.

Here’s what’s happening:

Microsoft’s documentation may be out-of-date or may not yet reflect current platform behavior.
In practice, Power Platform environments (and Azure connectors in general) sometimes allow higher timeouts due to backend changes or environment-specific configurations.
If 240 seconds works, the environment clearly accepts higher values, even if not officially documented.
The error for 300 seconds suggests there is an enforced cap, just higher than the docs say—possibly 240,000 ms (4 minutes) is the true current upper limit.

There’s no known official update confirming the documented max is above 30,000 ms, but your experimentation shows you can safely use up to 240,000 ms. To double-check, keep an eye on the product’s release notes or official Microsoft documentation for any policy clarifications.

Why the difference?

Microsoft often rolls out backend changes before documentation catches up.
Certain connector framework internals may permit higher limits unless strict policies are enforced at the UI or environment level.

You should use up to 240,000 ms if it works reliably in your environment, but avoid “banking on” this behavior for production or multi-tenant deployments until Microsoft confirms/supports it in docs or support channels.

Hope this clears things up!

Best regards,
Jerald Felix

Quick Links