← Back to search

Vibe Coding with Agents: Is the code any good? Which is best for you? Claude? Google? OpenAI?

Internet of Bugs 16:36

44,238 views · 1,998 likes Watch on YouTube ↗

Results of a round of comparison tests for the AI Coding Agents: Claude Code, GeminiCLI, and OpenAI Codex.

View detailed breakdown, code quality analysis, setup information on @SpecAgain at: https://youtu.be/Jg7Z8JYi6nU

Test your coding against the AIs and learn how the Internet works at https://IoB.FYI/CoCr

Info on potential for AIs to improve their own code:
https://www.forbes.com/sites/robtoews/2024/11/03/ai-that-can-invent-ai-is-coming-buckle-up/
https://www.youtube.com/watch?v=0bnxF9YfyFI

00:00 Introduction
01:04 Why these tests are important
02:22 What is "Vibe Coding"?
03:27 How I structured my testing
05:17 Summary Results
06:29 Setup Information
07:45 When "Vibe Coding" Agents go bad
09:02 Why I'm not using MCP
09:42 The One That Failed Miserably
11:28 The Two That Finished: How was their code?
15:11 Upcoming Plans for more Testing on this channel
15:56 Risks of "Vibe Coding"

Playback is via YouTube's official embedded player. Data from YouTube; Exumo is not affiliated with YouTube.