Observed arrival · 2026-10-13
NooBench’s six-app coding test
A side-by-side report compares Claude Pro and ChatGPT Plus on six one-prompt app-building briefs, tracking build time, reported limits, API cost, and runnable outputs.
- For
- Developers comparing AI coding-agent subscriptions
- Worth noticing
- Each app was built in its own sandbox; the page uses CLI-reported limits and labels the six-app comparison a first result, not a final verdict.
Field notes
The comparison uses Claude Code 2.1.296 with Opus 5.5 and Codex CLI 0.160.0 with GPT-6.1 Sol, each at high effort in a separate sandbox. The six briefs range from a 3D car showroom and a React Native habit app to a writer’s site; the page links to each prompt, app, and code. It reports limits as each CLI reports them and lists API prices at stated Oct. 10, 2026 list prices. There was one run per brief, with no follow-up prompts or fixes.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue