Response Streaming, Disclaimer Removal, and Token Limits in Copilot Studio

(1) Share

Report

Posted on by SG-17031007-0

Hi everyone, I have a few questions related to Microsoft Copilot Studio:

Response Streaming: Can we enable response streaming to a custom agent? If yes, how can it be configured? I couldn't find any documentation or settings for this.

Disclaimer Removal: For every response from the bot, a statement "AI generated content may be incorrect" is shown. Is there a way to remove or hide this disclaimer?

Token and Request Limits: What are the token limits for a single request/response? Are there any rate limits or request limits per minute/hour? Is there any documentation available for these limits?

Any guidance or documentation on these topics would be helpful. Thank you!

Categories:

General topics

I have the same question (0)

All responses (4)

Answers (0)

Suggested answer

Vinoth Selvam 1,590 Super User 2025 Season 2 on at

Like (1)

Report

Hi SG-17031007-0,

1. Response streaming meaning, what exactly you are expecting? Currently, Microsoft Copilot Studio does not support response streaming for custom agents. All responses are sent as complete messages once processing is finished.

2. The disclaimer "AI generated content may be incorrect" is automatically appended to bot responses to inform users about the potential inaccuracies inherent in AI-generated content. As of now, there isn't a supported method to remove or hide this disclaimer.

3. The default quota allows up to 8,000 requests per minute (RPM) per Dataverse environment. For standard subscriptions, there is a limit of 250,000 requests every 24 hours.

Reference link: https://learn.microsoft.com/en-us/microsoft-copilot-studio/requirements-quotas?utm_source=chatgpt.com

Thanks.

Visit my Blog

Was this reply helpful? Yes No
SG-17031007-0 4 on at

Like (0)

Report

@Vinoth Selvam, thank you for your prompt response. Also, do custom agents support caching functionality (like responses of already asked questions get stored somewhere for faster responses)?

Was this reply helpful? Yes No
Suggested answer

Vinoth Selvam 1,590 Super User 2025 Season 2 on at

Like (0)

Report

Hi SG-17031007-0,

Microsoft Copilot Studio does not natively support caching functionality for storing responses of already asked questions for faster retrieval. That too storing the values is definitely not happening.

Can i check what is the need for that. Do u feel the response might be faster if it stores?

Sometimes Copilot gives different responses for the same question if asked again. It is always the AI to learn based on our data and if we provide more context, the response will be more accurate. So caching is not going to help us in anyways.

Thanks.

Visit my Blog

Was this reply helpful? Yes No
SG-17031007-0 4 on at

Like (0)

Report

@Vinoth Selvam Got it, Thanks. Yes, I was thinking that responses would be faster

Was this reply helpful? Yes No