GPT-5 vs Claude Opus 4.1 vs Sonnet 4.5
Three-way comparison of AI coding capabilities
View on GitHubhttps://github.com/with-logic/gpt-5-vs-opus-4-1-coding-examples
Why This Comparison?
At Logic, our team starts every feature in Claude Code. Over the past 30 days, more than 85% of our code has been written by Claude. We've found incredible success with agentic coding, so naturally we were excited to compare how GPT-5 might improve upon the success we've already been having.
Each example below was generated using identical prompts given to OpenAI's GPT-5, Anthropic's Claude Opus 4.1, and Anthropic's Claude Sonnet 4.5. All applications were created in a single prompt without manual editing, showcasing the raw capabilities of each model.
Click on any example to view the implementations and compare their approaches to the same challenge.
