LOCAL INTELLIGENCE

Your work, in focus.

Scanning Codex sessions…

API-equivalent cost—Based on current standard pricing
Total tokens——
Cache hit rate——
Sessions / turns——
Uncached input——
Reasoning output——
Average session——
Busiest day——
Active days——
Models used——
Average model call——
Median active day——
Output intensity—Output as share of all tokens
Estimated cache saving—Versus standard input rates

MODEL TIMELINE

Which models powered each day

Stacked totals by model · hover any day for details

MODEL MIX

Models

COMPOSITION

Where tokens go

—tokens

WORKSPACES

Project overview

Click a bar to drill down
ProjectTokensInputOutputCacheEst. costTurns

DEEP DIVE

Heaviest sessions

Top 100 for current filter
StartedProjectModelTokensCachedOutputEst. cost