● Model Layer · I02 / I04 · Operational Telemetry

Model Token & Cost Accounting (I02 / I04)

Disjoint token components per model, plus observed and estimated cost kept apart. Reasoning stays inside output. Unknown cost is not zero. Task, actor, and model token accounting across invocations.

Model Layer Tracks physical 4-component token burn (uncached input, cache read, cache write, output) and prompt cache hit efficiency per model and actor without currency guesswork.
Total Recorded Tokens
14.24M
Physical token units · Zero inferred currency
Cache Read Fraction of Input (I02)
76.8%
10.94M read · Prompt cache reuse
Fresh Input (Uncached)
1.85M
13.0% of total tokens · New prompt tokens
Generated Output
1.45M
10.2% of total tokens · Model reasoning/edits
Task-Intake Overhead
3.8%
LOW OVERHEAD · Ambiguous/unlinked

4-Component Token Distribution by Model

Fresh uncached input vs cache reads vs cache writes vs generation output

Input Uncached
Cache Read
Cache Write
Generated Output
Task & Actor Token Burn Records 0 Rows
Actor ID Model Family Effort Attribution Bucket Input Uncached ▾ Cache Read Cache Write Output Total Tokens Cache Read Fraction of Input (I02)
Technical Lineage, Coverage & SQL Proof (L3)
Underlying SQL Projection
SELECT * FROM v_mission_token_burn ORDER BY input_uncached DESC, seat_id;