Genesys Casual

Hi everyone,

While reviewing the documentation for working with Topics in Speech & Text Analytics (specifically the "Generate phrase suggestions with AI" feature), I noticed that the article does not clearly mention how token consumption is handled for this functionality.

The feature allows us to generate up to 10 phrase suggestions per request using AI, which is great from a productivity standpoint. However, I could not find explicit information about:

• Whether this action consumes AI Experience tokens

• If token usage is calculated per request, per suggestion, or per character processed

• Whether the consumption model differs from other AI-powered features (like summaries, Copilot, etc.)

Since token governance and cost predictability are critical for enterprise environments, it would be helpful to understand how this specific AI-assisted topic creation is billed.

Could someone clarify:

• Does generating phrase suggestions for Topics consume tokens?

• If so, where can we find the official documentation detailing the billing model for this use case?

Thanks in advance - looking forward to better understanding how this works from a governance and cost management perspective.

#SpeechAndTextAnalytics
#AIExperience
#TokenBilling

#General

------------------------------
Mateus Nunes
Tech Leader Of CX at Solve4ME
Brazil
------------------------------

Hi Mateus,

There is no token consumption. There is a usage limit of 100 generations per day per org.

This is similar to the generate Intents and Utterances in Virtual Agent and Agent Copilot. The feature is effectively covered by the STA Add On, WEM Add On or CX3/4.

------------------------------
Richard Chandler
Connect
------------------------------

Original Message:
Sent: 03-02-2026 11:11
From: Mateus Nunes
Subject: How is AI token consumption calculated for Topic phrase generation?

Hi everyone,

While reviewing the documentation for working with Topics in Speech & Text Analytics (specifically the "Generate phrase suggestions with AI" feature), I noticed that the article does not clearly mention how token consumption is handled for this functionality.

The feature allows us to generate up to 10 phrase suggestions per request using AI, which is great from a productivity standpoint. However, I could not find explicit information about:

• Whether this action consumes AI Experience tokens

• If token usage is calculated per request, per suggestion, or per character processed

• Whether the consumption model differs from other AI-powered features (like summaries, Copilot, etc.)

Since token governance and cost predictability are critical for enterprise environments, it would be helpful to understand how this specific AI-assisted topic creation is billed.

Could someone clarify:

• Does generating phrase suggestions for Topics consume tokens?

• If so, where can we find the official documentation detailing the billing model for this use case?

Thanks in advance - looking forward to better understanding how this works from a governance and cost management perspective.

#SpeechAndTextAnalytics
#AIExperience
#TokenBilling

#General

------------------------------
Mateus Nunes
Tech Leader Of CX at Solve4ME
Brazil
------------------------------