Overview
Select a date range and fetch data
→
Token data comes from /user_usage_report and /usage_report. Engagement counts (messages, sessions) come from /users. Data is available 3 days after the usage date; the cost/usage endpoints refresh every ~4 hours.
Input Tokens (uncached)
–
across all users & products
Output Tokens
–
across all users & products
Cache Read Tokens
–
prompt cache hits
Cache Write Tokens
–
cache creation cost
Top Users by Total Tokens

No data yet

Fetch to populate
Token Type Breakdown
Fetch data
Token Usage Over Time (all users)
Org Model Mix
from usage_report · token share by model

No data yet

Fetch to populate
0 users from /user_usage_report · sorted by total tokens
#User InputOutputCache ReadCache Write Total RequestsWeb Searches
Fetch data to see per-user token breakdown
Daily Active Users (/summaries)
Seats Assigned
–
DAU (latest day)
–
Seat Utilization
–
DAU ÷ assigned seats
from /users endpoint · chat messages, Claude Code sessions, web searches
#User Chat MsgsConversations Code SessionsCommitsPRs Office SessionsWeb Searches
Fetch data to see engagement metrics
Range: – days
Est. Total API Cost
–
across all users
Avg Cost / User
–
for date range
Total Users
–
with token activity
Per-User Cost Breakdown by Model

Click "Load Model Breakdown" above

Fetches per-user, per-model token data to estimate API-equivalent costs
Compare Two Date Ranges
→ vs →

Select two date ranges to compare token usage

Export Raw Data

Download fetched data as CSV or JSON.

Raw API Response Inspector
Showing first record only. Check field names here if tokens appear as zero.
# Click a button above to inspect raw API response