Public usage ledger
AI Footprint
A public record of how AI participates in my creative and technical work. It shows aggregate usage, estimated cost trends, and model distribution without exposing conversation content.
At a glance
Estimated cost and aggregate token usage for the latest 30-day window.
Token trend
Daily changes in recorded token usage over the last 30 days.
Model ranking
The relative share of recorded usage across models.
Provider mix
The relative share of recorded usage across providers.
Activity & efficiency
Active days, sessions, average estimated cost, and cache efficiency.
How to read the data
The default window is the latest 30 days. Token counts are aggregated from usage fields recorded by supported tools, while cost is an estimate based on the recorded model and available pricing. Tool-specific logging, missing records, and pricing changes can introduce differences, so these numbers are best used to understand trends and composition rather than reconcile an invoice.
The public data boundary
This page publishes aggregate model, provider, token, estimated cost, and activity data. It does not publish prompts, responses, private project names, file paths, keys, or session details that could reconstruct a specific task. The purpose is to explain the structure of AI usage, not the content of the work.