Usage & Analytics Overview
The Usage page gives you visibility into AI model token consumption across your entire organization, broken down by model, project, department, and time.
What Is Tracked
Section titled “What Is Tracked”Every chat completion request processed by Menatic AI records:
- Timestamp — when the request was made
- Model — which model was called (e.g.
llama3:8b,gpt-4o) - Provider —
ollama,openai,gemini, orcustom - Prompt tokens — tokens in the input
- Completion tokens — tokens in the response
- Total tokens — sum of prompt and completion
- Project — which project the request was attributed to (if authenticated with a project token)
Usage is tracked for all requests including:
- Chat from the Menatic AI UI
- Requests via the OpenAI-compatible API using project tokens
- Notebook code using the
openaiSDK with the project’s API key
Usage Dashboard
Section titled “Usage Dashboard”The Usage page provides:
- Time-series chart — daily token consumption over a selected date range, broken down by model
- Hierarchy table — org → department → project drilldown showing cumulative token totals
- Date range selector — 24h, 7d, 30d, or 90d presets