Marathon debugging on meteo-brief
2.5M tokens vs typical 139.9K
66h 41m session running commands and testing. 15.3M tokens across 64 prompts using claude-opus-4-8[1m].
Derived from this session's token and cost data. Not shown on the feed.
Comments