HaiAI123

Curated Global AI Tools Directory

Industry update
Flat

Cost comparison of Codex CLI and Claude Code on the same tasks

10/06/2026 — 10/06, 18:17·1 sources (all)·1 reports (all)

Key takeaways

  • A group of developers ran four real-world tasks on a fixed open-source repository, twice per CLI in headless mode, to compare the costs of C…

Background & analysis

On October 6, 2026, a post on DEV Community published a head-to-head cost measurement of Codex CLI and Claude Code running the same tasks. The authors argue that most "Codex vs Claude Code" comparisons either quote vendor benchmarks or recycle a single third-party token figure that gets copied from page to page, and that few of them publish the tasks, the prompts, or the commit, which leaves their numbers impossible to verify.

They describe their own approach as the boring version. Four real tasks were run against one pinned open-source repository. Each CLI ran every task twice, in headless mode, with the exact commands and prompts set out alongside the results.

Pricing came from the official pricing pages on the day of measurement. The post states the measurements were taken on 2026-10-06 using Claude Code 2.1.291 with claude-opus-5-5.

So that others can check or rerun the work, the authors released the prompts, the hidden acceptance tests, the run scripts, and the JSON records of all 16 runs — four tasks, two CLIs, two runs each.

The available excerpt cuts off mid-sentence right after naming the Claude Code version. It does not include the Codex CLI version used, the cost figures for any of the four tasks, or any conclusion about which tool came out cheaper.

AI-generated from 1 reports · updated 1 day ago

Latest turnA team ran four real tasks on one pinned open-source repository, twice per CLI in headless mode, to compare Codex CLI and Claude Code costs. They published the prompts, hidden acceptance tests, run scripts and the full JSON transcripts of all 16 runs so the numbers can be checked or rerun.

Related AI tools

24-hour heatHeat index 36 · peak 71 · 23h ago
24 hours agonow

Reports on this story headlines open the original

Oct 6
  1. A team ran four real tasks on one pinned open-source repository, twice per CLI in headless mode, to compare Codex CLI and Claude Code costs. They published the prompts, hidden acceptance tests, run scripts and the full JSON transcripts of all 16 runs so the numbers can be checked or rerun.

    DEV Community · AIAI score 58

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →