One data agent for people and AI agents
Any agent.
Any database.
One governed answer.
Your team asks Wren's GenBI agent in plain language. Claude, ChatGPT or your own agents call Wren as a sub-agent over MCP. Same answer either way, from a context layer you own, across 20+ sources, on your servers or ours.
17,768stars on GitHubTrusted by data teams worldwide
Why Wren AI
Versioned in git. Shared by people and agents.
Not just a tool call. A sub-agent that reasons in a sandbox.
- Isolated sandbox
- Replayable traces
- Benchmarked SQL
Row- and column-level security at query time.
- UI, API, Slack and MCP
- Role-based access
- Full audit log
Integrations
- Snowflake
- BigQuery
- Databricks
- Redshift
- PostgreSQL
- SQL Server
- Oracle
- MySQL
- ClickHouse
- Athena
- Trino
- Starburst
- ChatGPTMCP
- ClaudeMCP
- GeminiMCP
- Slack and TeamsChat
- Your productEmbed
Any agent
ChatGPT, Claude, Gemini or your own agents call Wren as a sub-agent over MCP, and each one gets the same governed answer.
Any database
20+ sources, no ETL and no migration. Run it in our cloud, your VPC, or fully air-gapped.
Unlimited users on every paid plan
Open source and warehouse-agnostic, with pricing that doesn't grow with seats.
For leaders and their teams
Wren's GenBI agent turns one prompt into drill-down, roll-up and filtered dashboards on your governed data.
ASK WREN AI
|
Open core
Run it anywhere. Inspect every query. No vendor account required.
- mdl/orders.yml
- knowledge/metrics.md
- skills/q3-margin.md
Your model & knowledge, in your git
Case studies
Data shifted from an "expert tool" to a "company-wide asset."
For product & platform teams
iframe, API or MCP. Priced by usage, not seats.
- Signed JWTs
- Row & column-level scoping
- Full audit logs
- Per-app tokens
Start with your team, or your agents.
Try the GenBI agent in Wren AI Cloud, or talk to us about connecting your AI agents over MCP. Run it in our cloud, your VPC, or fully air-gapped on-prem.



