Verified comparative study of weather brief LLM models — 6.1M tokens

Verified comparative study of weather brief LLM models

meteo-brief

MARATHON

37.6x your typical session

6.1M tokens vs typical 161.3K

6.1Mtokens

32prompts

1:19:22time

88lines

Claude Codeclaude-opus-4-8[1m]Markdown$73.40

mostly delegating

Completed comparative evaluation of two LLM models with human-verified judge reviews and final scoring.

Tokens / prompt189.4k

Cost / line$0.834

Cache hit91%

Burn rate$55.49/hr

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

Building 14%Verifying 0%Thinking 0%Delegating 85%