Qonic Launches In-Model AI Agents and MCP Server for BIM Platform
AI assistants · Architecture · BIM collaboration
Qonic has launched Qonic AI, integrating agents directly inside its cloud-based collaborative BIM platform to handle data verification and geometry editing via natural language. The company also introduced an MCP server to connect external AI assistants.
BIM software developer Qonic has introduced Qonic AI, embedding agentic capabilities directly into its cloud-based collaborative modeling platform. Rather than operating as an external add-on, Qonic AI runs within the model environment, allowing users to find, check, and edit BIM data and geometry through the 'Ask Qonic' prompt interface. Capabilities include assigning missing properties, generating grouped reports, and resizing geometric elements. Alongside the in-platform agents, Qonic launched an MCP server that enables external AI assistants to query elements, compute quantities, and export models to IFC.
Key takeaways
- Qonic AI operates directly inside the BIM model, allowing users to query properties, correct missing data, and adjust geometry through text prompts.
- The system enforces permissions, restricts edits to authorized models, and allows all AI modifications to be undone like manual actions.
- Users can select underlying AI models, including Sonnet, Opus, or Astra, using a token-based billing structure with monthly allowances or standalone packs.
- A newly launched MCP server connects external assistants such as Claude or Mistral Le Chat to query elements, calculate quantities, and export IFC files.
Source：AEC Magazine Read the original ↗
Prepared, checked and automatically published by AI; not individually reviewed by a human.