Wren AI
Open-source governed text-to-SQL and GenBI layer that turns questions into trusted dashboards.
About
Wren AI is an open-source (Apache 2.0) generative-BI layer for AI agents: instead of letting an agent guess SQL against raw schemas, you model business definitions in versioned MDL semantic files, and Wren provides schema-aware retrieval, dry-plan validation, and structured errors so agents produce governed, correct SQL. It connects to 20+ data sources — BigQuery, Snowflake, PostgreSQL, ClickHouse, Redshift, Databricks — works through the agents you already use (Claude Code, Cursor, MCP), and can turn any answer into a shareable browser dashboard deployable to your own Vercel or Cloudflare Pages account.
The honest picture: the governance is only as good as the semantic models you write — MDL setup is real upfront work, and a thin model gives thin results. Row/column-level security lives in Cloud/self-hosted tiers, not the pure OSS core, and it is overkill if you just need a one-off chart from a CSV.
What it can do
- Governed text-to-SQL with dry-plan validation
- Semantic layer (MDL) versioned in Git
- Agent-driven dashboard generation and deployment
- 20+ data sources
- Works via Claude Code, Cursor, and MCP
- Apache 2.0 open-source core
Setup
Deploy the open-source stack via Docker or use Wren Cloud; define semantic models in MDL files.
Prerequisites
A data warehouse connection; willingness to model business semantics in MDL.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 50; Trending score 74/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.