Marathon debugging on frontend
2.6M tokens vs typical 414.6K
67h 11m session running commands and testing. 2.6M tokens across 53 prompts using claude-haiku-4-5-20251001.
Derived from this session's token and cost data. Not shown on the feed.
Comments