Message 1 of 5
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report
AutoCAD exposes a very large number of APIs, which are extremely detailed and technical. I am exploring how Large Language Models (LLMs) can be used to assist with AutoCAD automation, but expecting an LLM to understand and safely work with the entire AutoCAD API surface for every user request is complex, expensive in terms of tokens, and prone to hallucinated or incorrect API calls. With a pure CS/IT background (no mechanical or civil domain), I am looking for guidance for the same. I would appreciate suggestions, best practices, or architectural patterns from the community on how to design such a system in a reliable, scalable, and production-safe way.
Solved! Go to Solution.
