Me:: Simon's recommended models as of 2025-03-02: Claude 3.7 Sonnet w/'thinking' on, OpenAI o3-mini-high & GPT-4o with Python Code interpreter Simon Willison:: Hallucinations in code are the least dangerous form of LLM mistakes

Discovered: Mar 2, 2025 06:56 Me:: Simon’s recommended models as of 2025-03-02: Claude 3.7 Sonnet w/’thinking’ on, OpenAI o3-mini-high & GPT-4o with Python Code interpreter Simon Willison:: Hallucinations in code are the least dangerous form of LLM mistakes <– QUOTE: Try different models. It might be that another model has better training data for your chosen platform. As a Python and JavaScript programmer my favorite models right now are Claude 3.7 Sonnet with thinking turned on, OpenAI’s o3-mini-high and GPT-4o with Code Interpreter (for Python).