ILLUSTRATION ONLY — every figure below is randomly generated sample data. No backend, no accounts, no real usage is tracked.
spend this period
$1,842.60
↑ 12% vs. prior period
tokens processed
742.3M
↑ 8% vs. prior period
avg. cost / 1k requests
$3.11
↓ 4% vs. prior period
active projects
6
2 new this period
spend over timedaily, USD
spend by provider
monthly budget$2,000 cap
$1,842.60 of $2,000 used (92%) · alert threshold set at 80% · projected to exceed cap in ~3 days at current pace
spend by team / project
project
tokens
requests
spend
% of total
requests today
48,120
↓ 2% vs. yesterday
error rate
0.34%
↓ 0.1pt vs. yesterday
p95 latency
184ms
↑ 6ms vs. yesterday
uptime (30d)
99.97%
2 incidents
request volumehourly, last 24h
key rate-limit status
api keys
name
created
last used
status
recent requests
time
endpoint
model
tokens
latency
status
Sample data regenerates on page load and is not stored anywhere. Wiring this to real data requires
auth (so each org only sees its own usage), a database logging every API call, and a usage-tracking
layer in api/*.js — see README for the outline when you're ready to build that.