Debugged que-fer — 20.5M tokens

Debugged que-fer

meteo-brief

DEEP WORK

127.2x your typical session

20.5M tokens vs typical 161.0K

20.5Mtokens

34prompts

26:14:55time

620lines

Claude Codeclaude-opus-4-8[1m]$367.77

mostly building

84h 42m session running commands and testing. 19.4M tokens across 23 prompts using claude-opus-4-8.

Tokens / prompt602.4k

Cost / line$0.593

Cache hit96%

Burn rate$14.01/hr

Derived from this session's token and cost data. Not shown on the feed.

Building 66%Verifying 24%Thinking 9%Delegating 2%