Share via

Billing issue with safe evaluations input tokens in AI Foundry

Sam Wynn 15 Reputation points
2025-07-17T17:35:22.31+00:00

We have a few simple agents that we only use in the chat playground. Talking to the model itself has cost us around 2 dollars in chatting back and forth. When looking at our billing we are seeing this:User's image I have been looking for the option to disable this as its a 1000% increase in the price. Do you know where this feature is?

Azure AI Content Safety
Azure AI Content Safety
An Azure service that enables users to identify content that is potentially offensive, risky, or otherwise undesirable. Previously known as Azure Content Moderator.
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Markus Hofer 0 Reputation points
    2025-07-30T08:17:27.4933333+00:00

    This is a complete dealbreaker! I experienced the same issue. I also did not find a way to disable this safety evaluations.

    In my case I have an AI agent which uses Grounding with Bing and GPT 4.1. Grounding with Bing is already extremely expensive (USD 35/1000 transactions). But the hidden cost caused by the AI agent are even twice as high.

    In my case the cost for the LLM is completely irrelevant compared to the additional cost caused by the pure Microsoft services. In my case this means that the Azure AI Agent cannot be used for the project (which could even mean that Azure will not be used at all).

    Service % of total cost
    LLM (GPT 4.1) 2.5 %
    Grounding with Bing 30.3 %
    Agent - Safety Evaluations Input/Output tokens (hidden cost) 67.2 %

    Microsoft should explain why those hidden cost apply and how they can be avoided.

    0 comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.