LLM Digest

Story

hackernews_ai · May 21, 2026 · news

Source brief

I tested Haiku vs. Sonnet across 3 agent tasks – the cheap model won every time

github.comMay 21, 2026
original source linked

In brief

Article URL: https://github.com/aimvik07/agent-eval Comments URL: https://news.ycombinator.com/item?id=48225730 Points: 3 # Comments: 0