GPT-4o breaks code structure under strict prompts (compared to GPT-4-turbo)

This repository documents specific regressions experienced when using GPT-4o compared to GPT-4-turbo.

To help OpenAI and the community track behavioral changes in GPT-4o that impact technical work, especially for users relying on strict, repeatable code generation workflows.

• GPT-4o_Feedback_Report.md: Full markdown report sent to OpenAI Support.
• examples/: Individual breakdowns of regression cases.
• .github/ISSUE_TEMPLATE.md: Optional issue format for contributors.

Created and maintained by Ryan (@chapman4444)

🗣️ Join the discussion on Reddit