---
schema_version: '1.0'
id: news-20260713-398431
url: https://systima.ai/blog/claude-code-vs-opencode-token-overhead
url_hash: 398431a5f118d841d6a8dd041caabbd674f935908d0bf2ed36b173c37cd5c111
canonical_url: https://systima.ai/blog/claude-code-vs-opencode-token-overhead
source: HackerNews
category: news/tech
category_raw: tech
region: usa
tags: []
lang: en
published_at: '2026-07-12T18:25:51Z'
fetched_at: '2026-07-13T00:28:36Z'
updated_at: '2026-07-13T00:28:47Z'
status: published
content_hash: null
license_note: link-only
summary: null
summary_source: null
summary_en: null
entities:
- name: Claude
  type: creature
- name: Codeberg
  type: organization
- name: THE ANSWER
  type: organization
- name: Prompt Compression Technology
  type: method
key_facts: []
related: []
related_auto:
- name: Context Pack
  type: concept
  weight: 2.0
- name: AI時代
  type: concept
  weight: 1.0
- name: Anthropic Natural Language Autoencoders
  type: event
  weight: 1.0
- name: 会話資産
  type: concept
  weight: 1.0
- name: techcrunch
  type: organization
  weight: 1.0
title: Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k
---

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

## TL;DR
（要約なし・原文はリンク先を参照 / No summary available; see the source link.）

## Key Points
- (なし)

## Details
(本文なし。リンク先参照)

## Source
元記事: [Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k](https://systima.ai/blog/claude-code-vs-opencode-token-overhead) — published 2026-07-12T18:25:51Z