Wren AI

Overview

Wren AI is a Generative BI (GenBI) platform that lets you query your databases using plain English instead of SQL. It connects to 20+ data sources like BigQuery, Snowflake, PostgreSQL, and Databricks, using a semantic layer (MDL) to turn questions into trusted SQL, charts, and dashboards. Built for data teams, business users, and AI agents, it works as a standalone app, an embedded API, or an MCP server for Claude, Cursor, and Slack. Available as open-source or commercial cloud/on-prem editions.

Key Features

Natural Language to SQL: Ask questions in plain English and get accurate SQL queries with validation against your database to reduce hallucinations.
Semantic Layer (MDL): Modeling Definition Language captures metrics, relationships, and business logic in version-controlled files that every consumer resolves through.
20+ Data Source Connectors: Native connections to BigQuery, Snowflake, PostgreSQL, ClickHouse, Amazon Redshift, Databricks, Oracle, and more without rebuilding pipelines.
Real-Time Dashboards & Charts: Generate interactive dashboards with drill-down, roll-up, and global filters directly from natural language questions.
Embedded API: Add conversational analytics to your own SaaS product with a single API call that returns SQL, data, and ready-to-use charts.
MCP & Agent Integration: Drop Wren AI into Claude, Cursor, ChatGPT, Slack, or Teams as a governed context layer for AI agents.
Knowledge & Memory: Save approved definitions and past interactions as reusable context so answers stay consistent over time.
Flexible Deployment: Choose managed cloud, private cloud, or air-gapped on-premises for full data control and compliance.

Pros

Open-Source Foundation: Apache-2.0 licensed core with active Discord community and weekly updates gives you transparency and flexibility.
Broad Data Source Support: Works with 20+ databases and dbt sync, so you can use your existing stack without migration.
Trusted Text-to-SQL Generation: Uses semantic modeling, context retrieval, SQL validation, and query planning to generate more reliable answers from business data instead of relying only on raw database schemas.
Customizable Knowledge and Context Layer: Allows teams to add business definitions, metadata, verified query examples, and instructions so AI responses match company-specific terminology and rules.
Semantic Modeling with MDL: Uses Modeling Definition Language (MDL) to define data relationships, metrics, calculations, and business logic, helping both users and AI agents understand company data.
Agent-Ready Architecture: Supports MCP and integrations with AI agents such as Claude and Cursor, allowing agents to access the same governed business context layer used for analytics.
AI-Generated Analytics and Dashboards: Converts natural language questions into SQL answers, charts, and dashboards, making data exploration easier for business users.

Cons

Self-Hosting Requires Technical Knowledge: Deploying and managing Wren AI within your own infrastructure may require experience with databases, LLM configuration, and system administration.
Learning Curve for Semantic Modeling: Setting up Modeling Definition Language (MDL), defining relationships, and creating business logic requires time and data modeling expertise, particularly for teams building advanced analytics workflows.
Commercial Pricing Requires Careful Planning: Self-hosted commercial plans utilize concurrent session licensing rather than named-user pricing; therefore, organizations should estimate their usage patterns before selecting a plan.
AI-Generated Results Still Require Review: Although Wren AI utilizes a governed context layer, semantic models, and validation workflows to improve accuracy, users should still verify important business queries and reports.

Use Cases

Self-Service Analytics for Business Teams: Marketers, sales managers, and operations staff can ask questions like "revenue by product category last quarter" without waiting for the data team. This removes bottlenecks and speeds up decisions.
Embedded Analytics in SaaS Products: Product teams use the Embedded API to let their customers query their own data with natural language inside the app, complete with custom branding and governance.
Executive Reporting & Strategy: CEOs, CROs, and CMOs run expansion analysis, revenue lift studies, and board memos by asking direct questions instead of commissioning reports.
HR Data Queries: HR teams get answers about headcount, attrition, and compensation without needing analyst support for complex SQL joins.
AI Agent Data Access: Developers connect Claude, Cursor, or custom agents to enterprise data through the MCP server, giving agents a governed way to query business data.
Automated Recurring Reports: Data teams offload standard reporting to Wren AI dashboards, reclaiming hours each month for higher-value work.

Frequently Asked Questions

Is Wren AI open source?

Yes. Wren AI OSS is an open-source GenBI platform that users can self-host and customize. The open-source edition provides the core context layer, semantic modeling capabilities, and developer tools, while commercial editions add features such as team management, security controls, integrations, and support.

Which databases does Wren AI support?

Wren AI connects to 20+ data sources including BigQuery, Snowflake, PostgreSQL, ClickHouse, Amazon Redshift, Databricks, Oracle, and file sources. It also supports dbt sync for teams already using dbt models.

How does Wren AI reduce SQL hallucinations?

It combines a semantic layer (MDL), rich schema retrieval, and dry-run validation against your database. If a generated query fails, the system uses the error feedback to regenerate a corrected version.

Can I use Wren AI with AI agents like Claude or Cursor?

Yes. Wren AI provides an MCP server so you can connect it to Claude, Cursor, ChatGPT, Slack, Teams, or any MCP-compatible client as a governed data access layer.

How is Wren AI priced?

Free plan: $0/month with 20 monthly credits. It includes basic GenBI features, visualization, knowledge base access, up to 2 projects, 2 members, and limits on tables per project.
Essential plan: $179/month when billed annually. It includes 13,200 annual credits, unlimited dashboards, unlimited spreadsheets, unlimited connectors, Embedded AI API access, dbt integration, API management, basic role-based access control, permission controls, and standard support.
Enterprise plan: $559/month when billed annually. It includes 24,000 annual credits, advanced security features, row-level and column-level data controls, MCP integration, unlimited integrations, advanced RBAC, audit logs, LDAP/AD support, OIDC support, and priority support.

Can I self-host Wren AI?

Yes. You can run Wren AI OSS on your own infrastructure, deploy the commercial edition in your private cloud, or use an air-gapped on-premises setup for regulated environments.

Do I need SQL skills to use Wren AI?

No. Business users can ask questions in plain English. However, technical users benefit from setting up the semantic model and knowledge base for the most accurate results.

Can I embed Wren AI in my own product?

Yes. The Embedded AI API lets you add natural-language SQL, data, and chart generation to your application with your own branding, governance, and data isolation.

Product information

What is Wren AI?

Categories

Social

Pricing

Overview

Key Features

Pros

Cons

Use Cases