Hacker News · item history
Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k
2
first seen points
701
peak points
701
latest points
35
observations
Posted on X
- Claude Code sends 33k tokens before it reads your prompt. OpenCode sends 7k. That's 4.7x the overhead — before the model does anything useful. Token efficiency isn't a nice-to-have. It's latency and cost, billed per request.