Onshape has steadily expanded its built-in AI capabilities, but the Leo AI article argues that engineers should distinguish between AI that helps operate the CAD system and tools that can access actual design data. Five capabilities are currently available: AI Advisor, LLM-powered FeatureScript generation, AI model search, AI Quick Render, and Replicate Annotations.
AI Advisor provides real-time guidance based on Onshape documentation and Learning Center resources. Engineers can ask about workflows, troubleshooting, modeling practices, mates, drawings, and other software-related topics. However, the assistant does not reason over the active model or understand its revision history, assemblies, or release status.
FeatureScript generation reaches closer to geometry. Onshape uses AI to generate and autocomplete FeatureScript code for custom parametric features. Its FeatureScript MCP Server also connects external language models with FeatureScript, allowing natural-language instructions to become reusable CAD features. The article stresses that AI-generated automation still requires engineering review.
Other AI functions address narrower tasks. AI-powered search uses generated model descriptions instead of relying only on filenames and metadata. AI Quick Render creates presentation images from text prompts, while Replicate Annotations uses machine learning to transfer dimensions, tolerances, layouts, and other annotations between drawings.
For deeper access to engineering information, Onshape’s REST API remains important. It exposes document structures, workspaces, versions, assembly relationships, release states, and permissions. This programmatic access can support AI systems that need to analyze actual project data across multiple documents.
Still, neither built-in AI nor the API automatically reaches information stored outside Onshape. Engineering knowledge may remain scattered across PDM and PLM systems, ERP platforms, standards, spreadsheets, and network drives.
The article presents Onshape’s AI strategy as a layered approach: built-in tools assist with specific CAD tasks, the API provides structured access to design information, and broader AI systems can connect knowledge distributed across engineering platforms.
