Power BI Custom Connector LLM Usage Monitoring Beta connector Free

LLM Usage Monitor (Power BI Custom Connector)

A beta Power BI custom connector for LLM usage monitoring that helps analysts track organization-level token usage, model consumption, request volume, and cost across OpenAI and Anthropic admin APIs.

Download .mez

What it does

A Power BI connector for the token and cost data behind AI usage.

LLM Usage Monitor is a beta connector that brings organization-level LLM usage APIs into Power BI as clean tables for usage, costs, models, API keys, projects, and service tiers. It is designed for analysts who need to explain where AI consumption is coming from, which models drive cost, and how usage changes over time.

Features

Designed to make LLM usage visible in Power BI.

Daily Usage and Cost Tables

Exposes daily token usage, request volume, and cost as typed Power BI tables.

Gives analysts a reliable model for tracking LLM usage without manually exporting billing screens or API reports.

Provider Selection

Supports OpenAI organization admin APIs and Anthropic organization admin APIs from one connector entry point.

Lets Power BI reports compare usage patterns across supported LLM providers while keeping provider-specific API details behind the connector.

Model-Level Visibility

Breaks usage down by model with input tokens, output tokens, total tokens, and requests.

Helps teams see which models are driving adoption and cost instead of treating LLM spend as one blended number.

API Key, Project, and Service Tier Views

Provides separate tables for usage by API key, provider project, and service tier when the provider API returns those dimensions.

Supports chargeback, operational review, and governance workflows for teams managing shared AI infrastructure.

Power BI Credential Store Authentication

Uses Power BI Key authentication so provider admin keys are entered through the credential prompt instead of stored in source files.

Keeps sensitive credentials out of reports, connector code, and documentation.

Refresh-Oriented Connector Design

Uses a stable data source kind, static base URLs, relative paths, typed schemas, and retry handling for common transient API failures.

Makes the connector easier to operate in Power BI workflows that need repeatable refresh behavior.

Connector inputs

Inputs and tables aligned to LLM operations.

Provider

OpenAI or Anthropic

Select OpenAI (Organization Admin API) or Anthropic (Organization Admin API only) from the connector prompt.

Admin API Key

Required

Use an organization admin key with access to usage and cost reporting APIs. Standard runtime API keys are not enough.

Usage Tables

6 tables

Load Usage, Costs, Models, UsageByApiKey, UsageByProject, and UsageByServiceTier into Power BI.

How to use

Install the connector, choose a provider, then load usage tables.

  1. 1 Download the LLMUsageMonitor.mez connector package.
  2. 2 Copy the .mez file into Documents\Microsoft Power BI Desktop\Custom Connectors.
  3. 3 Restart Power BI Desktop and enable custom connectors under Security options if required.
  4. 4 Open Get Data, search for LLM Token Usage, and select the connector.
  5. 5 Choose OpenAI or Anthropic as the provider and paste the matching organization Admin API key.
  6. 6 Load the usage, cost, model, API key, project, and service tier tables needed for your report.

Security Signal

Aikido logo

LLMUsageConnector

Repository security checks

0 issues Scanned version: Beta connector
Check Description Compliance Issues
Open source dependency monitoring
Monitors third-party dependencies and lockfiles for known vulnerabilities.
Compliant
0
Exposed secrets monitoring
Checks whether secrets have been accidentally exposed in the source repository.
Compliant
0
License management
Reviews dependency licenses for usage and compliance concerns.
Compliant
0
SAST
Runs static application security testing on the connector code.
Compliant
0
Infrastructure as Code
Checks infrastructure-as-code files where monitored files are present.
Compliant
0
Supply-chain malware
Checks dependencies for known malicious package signals.
Compliant
0
Mobile issues
Checks mobile manifest issues where monitored mobile files are present.
Compliant
0
GitHub access control issues
Monitors repository access and source-code security controls.
Compliant
0

The LLMUsageConnector repository scan reported 0 issues across monitored security checks.

This scan is a helpful third-party signal for the repository, but it does not replace your organization's own review of beta software, connector behavior, and credential handling.

Use cases

Built for teams responsible for AI usage visibility.

Data Analysts
AI Operations
BI Developers
Finance Operations
Platform Teams

Get the connector

Start monitoring LLM usage from Power BI.

Download the free beta LLM Usage Monitor .mez connector and follow the documentation to add it to your Power BI custom connector folder.