Overview
Per day, by tool
By device
Tool mix
Model mix
share of the selected range, and the tools that ran each
Project mix
ranked by the selected measure · click one for its sessions
Highlights
When you work
local hour × weekday, by tokens
Token mix
what each tool's tokens were spent on
Activity
last 12 months · click a day to zoom in
Monthly rhythm
tokens by month · click a month to zoom in
Cumulative spend
running total by tool · dashed = previous period
Spend by model
Spend by project
Spend by token type
what each kind of token cost you
Prompt-cache hit rate
cache reads as a share of context tokens
Effective rate by model
cost per 1M tokens
Daily spend
stacked by tool
Subscription limits remaining
Account-wide limits · tools used today on this device
Models
click a header to sort · a model used in several tools expands into its per-tool split
Model efficiency
how often each model's edits land first time, and what a prompt costs on it · Claude Code & Codex
Model timeline
top 8 models per day
Provider share
100% of each day
Provider × tool
where each maker's models actually ran
Top tool calls
dot = the tool that called it most
What the agents do
tool calls by category
What your prompts were for
each prompt and the work it set off
Edits that landed first time
share of editing prompts with no edit → run → re-edit of the same file
Where you work
which IDE or surface each tool ran in
MCP servers
calls, and tools used out of those offered
Skills
requests a Skill drove (Claude Code)
All tool calls
click a header to sort
Projects
by working directory
Project detail
click a row to see its sessions
Sessions
click a row for detail
Log bytes by tool
live files on disk right now
Bytes per 1M tokens
how expensively each tool stores its history
Storage over time
cumulative bytes by the day each log was last written
Largest logs
the files actually taking the space
Related data not analysed
AI data on disk that isn't turned into analytics