Wren AI is an open-source generative business intelligence platform from Canner, with nearly 18,000 stars on GitHub. It lets people and AI agents ask questions of company data in natural language and get back governed SQL, charts, and dashboards. What sets it apart is its context engine: a semantic layer that teaches LLMs your business definitions, so answers are consistent and trustworthy instead of improvised. The project releases weekly and is actively developed in 2026.
Wren AI models your data in MDL — a modeling definition language stored as code in git — capturing metrics, relationships, and business logic once, then applies those definitions at query time for every consumer. The open-source engine connects to more than 20 data sources, including BigQuery, Snowflake, PostgreSQL, ClickHouse, Databricks, and Redshift, querying data in place rather than copying it. Agents reach the engine over MCP, and the same core powers everything from the free CLI to the commercial cloud. Commercial plans add the GenBI app experience: dashboards, agentic analysis mode, AI-assisted context preparation, embedded APIs, and enterprise access controls.
The Wren AI context engine is free and open source, including the CLI and MCP server. Paid Cloud and Enterprise Plus plans add the full GenBI application, dashboards, and enterprise security and support.
Wren AI is for data teams and businesses that want reliable natural-language analytics — and agent access to data — without sacrificing governance, definitions, or control of their warehouse.