Back to feed
Dev.to
Dev.to
7/14/2026
The user wants me to rewrite a news headline for a mobile feed. Let me analyze the original:

The user wants me to rewrite a news headline for a mobile feed. Let me analyze the original:

Original: Claude Code burns 5x more tokens before you type a word. Here's where they go.

Short summary

A logging proxy study by Systima reveals Claude Code burns ~33k tokens before user input vs OpenCode's ~7k, due to 27 tool schemas, injected reminder blocks, and a larger system prompt. Subagent fan-out multiplies session cost by 4.2x, and unstable cache prefixes cause expensive mid-session rewrites. The harness, config, and subagent usage — not the model — are the real cost drivers in production AI coding setups.

  • •Claude Code opens sessions with ~33k tokens vs OpenCode's ~7k before any user input
  • •Subagent fan-out can silently 4x session token costs; 72KB instruction files add 20k tokens per request
  • •Instrument at the API boundary, not the dashboard, to find where tokens actually go

Generated with AI, which can make mistakes.

#ai-tools#ai-agents#open-source
Read full article at Dev.to

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

Study finds Claude Code sends ~33K tokens before reading user prompt, 4.7x more than OpenCode
Dev.to
Claude Code Sends 33k Tokens Before Your Prompt; OpenCode Sends 7k
Dev.to
The 33k-token preamble is a bet. Here's how to check if it's paying off.
Dev.to
The original title is about Claude Code's token usage and how a developer cut their bill in half using a tool called Lynkr.
Dev.to