GPT-5 Codex CLI vs Claude Code CLI – Full Head-to-Head Coding Test
Автор: Bijan Bowen
Загружено: 2025-09-17
Просмотров: 21131
Timestamps:
00:00 - Intro
01:00 - Test Methodology
01:40 - Testing Setup
03:16 - Passionate Individual Disclaimer
03:50 - Web Browser OS Test
05:15 - Claude Snooty Behavior
06:32 - Pre-Result Impressions
07:15 - Codex Web OS Result
09:22 - Claude Code Web OS Result
12:53 - Web OS Result Overview
13:49 - Second Test Setup
14:36 - Low Level Programming Test
16:18 - Testing Begins
18:36 - Pre-Result Impressions
21:41 - Improving Result Test
24:00 - Nothing Working
24:45 - Code Improvement Testing
25:45 - Test Difficulty High
26:50 - Final Chance Testing
28:00 - Why This Test?
28:41 - Closing Thoughts
AI Integration & Consulting: https://bijanbowen.com
Join the Discord: / discord
In this video, we put two of the most capable agentic coding tools to the test — the newly updated GPT-5 Codex CLI from OpenAI and Claude Code CLI from Anthropic.
We begin with a simple UI/browser-oriented test to showcase design reasoning, then push both models into advanced territory: attempting to write a low-level embedded driver to turn a Raspberry Pi Pico into an HID device capable of emulating a keyboard. Throughout the process, we evaluate planning, code quality, error recovery, and reasoning under failure.
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
![What One Person Can Ship with AI in 2025 | Claude Code Live Session [Extended Cut]](https://image.4k-video.ru/id-video/fAYkcwo45Ww)
