Work faster with the AI Assistant
Chat with an assistant that knows semantic modeling.
Build and change your model through chat, without leaving Tabular Editor's visual interface – so you can see and control everything the assistant does. It's powered by instructions from semantic modeling experts and runs on the LLM your organization already uses, including local models.
- Explore your model and write DAX queries
- Apply model changes directly from chat
- Run Best Practice Analyzer checks and fix issues
- Add custom instructions for your own standards
Bring your own agent with the MCP Server
Your agent does the work – you keep the final say.
Already using Claude Code, GitHub Copilot, Codex, or Cursor? Connect it to the model you have open in Tabular Editor – no extra API key needed. Every change shows up in your session, ready for you to inspect, undo, or save.
- Connect the agent you already use
- Review every change visually, right where you work
- Let the agent check its work with BPA and VertiPaq Analyzer
See and control every change
Full visibility, whichever way you work with AI.
With most AI tools, it's hard to tell what an agent actually changed. In Tabular Editor, every change is visible in the interface – so you keep what works and undo what doesn't.
- See changed objects and properties at a glance
- Revert selected edits – or everything at once
- Nothing outside your model runs without your review
- Agents check their work against your Best Practice Analyzer rules
Enterprise-ready AI
Set the boundaries once – for chat and agents alike.
The AI Assistant and the MCP Server share the same permissions, so you decide once what AI can access and do – and every action is logged.
- Allow model edits without granting access to your data
- Central governance through policies (Enterprise)
- Local audit log of all AI activity (Enterprise)
- Independently penetration-tested