Token and request monitoring
Track input tokens, output tokens, total tokens, and request volume by day so teams can see how LLM usage is changing over time.
Beta documentation
A beta Power BI custom connector for analysts who need to monitor organization-level LLM token usage, request volume, model consumption, and cost.
Connector output
Usage
Costs
Models
UsageByApiKey
UsageByProject
UsageByServiceTier Use case
Track input tokens, output tokens, total tokens, and request volume by day so teams can see how LLM usage is changing over time.
Bring daily cost data into Power BI so analysts and finance operations teams can review spend patterns without manually collecting reports.
Break usage down by model, API key, provider project, and service tier where the provider API makes those dimensions available.
Install
Connect
Provider requirements
Requires an OpenAI Admin API key with access to organization usage and cost endpoints. Runtime project keys are not sufficient for organization usage reporting.
Requires an Anthropic organization account, an admin role, and an Admin API key that starts with sk-ant-admin. Standard sk-ant-api keys and individual Anthropic accounts cannot access usage and cost reports.
Tables
Daily provider, date, total tokens, input tokens, output tokens, and requests.
Daily provider, date, and cost in USD.
Daily usage by model, including token and request measures.
Daily usage by provider API key identifier.
Daily usage by provider project or workspace identifier.
Daily usage by service tier when returned by the provider API.
Need help?
Reach out if you have questions about installation, provider admin keys, or how to model token usage and cost in Power BI.