Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k

Per Hacker News, Systima's technical analysis compares AI coding assistants on token efficiency. Claude Code sends thirty-three thousand tokens before even reading a user's prompt, while OpenCode sends just seven thousand. The study, which logged requests to Anthropic's endpoint, found Claude Code significantly less efficient in its cache strategy and overall token usage.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton