I'm on a mission to document my journey of becoming an AI-native founder, sharing every powerful workflow and hard-won insight along the way. You're welcome to join my newsletter ( ) or subscribe here, if you're a founder or software engineer looking to build faster and smarter with AI Agents like Claude Code, Codex, Cursor, Gemini, ChatGPT, n8n, image/video gen, and other AI workflows. The speech-to-text voice tool I use is called superwhisper. If you check it out via my affiliate link, it helps support the channel (I may earn a small commission): The simultaneous releases of OpenAI's Codex and Anthropic’s Claude Code 2.0 (powered by the new Sonnet 4.5) have completely overhauled my AI coding strategy. These two models are now battling for dominance, and the best workflow might not be choosing one-but learning how to use both effectively. This video is a CTO's real-world stress test, comparing Codex and Claude Code 2.0 across nine critical dimensions while migrating a massive legacy application. I break down the strengths and weaknesses of both—from Sonnet 4.5's incredible speed and "grittiness" to Codex's architectural reasoning and the revolutionary Codex Cloud platform. Finally, I share the exact hybrid workflow I’m using now, leveraging the unique capabilities of both models to maximize development velocity and code quality. Core Sections & Timestamps: 00:00 Introduction: Codex vs. Claude Code 2.0 (Sonnet 4.5) 01:18 1) Model Capabilities: Sonnet 4.5's Benchmarks, Speed, and "Grittiness" 05:48 2) CLI (Agent Harness): Claude Code 2.0 vs. Codex CLI 07:57 3) IDE Integration (VS Code & Cursor) 09:20 4) Codex Cloud: The New Paradigm of Cloud Agents 11:50 5) GitHub Agent & Reviewer: Automated PR Reviews and @Mentions 13:30 6) Vision & Future Roadmap: OpenAI's AGI Goal vs. Anthropic's Tactical Approach 15:35 7) Codex & Claude Code Pro Tips 16:45 8) Bonus Tip / Workflow: Using Gemini Deep Think for Research & PRDs 18:25 9) My New Workflow (and Model): Why I Switched to Sonnet 4.5 19:00 The Ideal Workflow: Codex Cloud for Research, Sonnet 4.5 for Implementation 19:33 Cross-Model Code Reviews 19:50 Model Personalities: The Entrepreneur (Sonnet 4.5) vs. The Academic (Codex) 21:00 Conclusion: The Importance of Orchestration Frameworks Links Mentioned in Video: Cognition report on context awareness: My video on Anthropic's Open Source Agentic Workflows (Code Review): My video on turning Claude Code into an amazing designer using Playwright: More About Me: Twitter: Personal Website: Thanks for watching! #Codex #ClaudeCode #Sonnet45 #AICoding #SoftwareEngineering #AgenticWorkflows #Anthropic #OpenAI #DeveloperProductivity #AIStrategy #CTO #GPT5