Vanna

Overview

Vanna is an open-source Python framework that turns natural-language questions into SQL queries for your own database. It uses retrieval-augmented generation trained on your schema, documentation, and past queries, and works with many LLMs and databases. The framework is free under the MIT license, with an optional hosted service.

Key features

• Natural-language to SQL over your own schema
• RAG-based training on DDL, docs, and example queries
• Works with many LLM providers and databases
• Python package plus optional web and Slack interfaces

Use cases

• Self-service analytics for non-SQL users
• Embedding text-to-SQL in internal tools
• Accelerating ad-hoc data exploration for analysts

Pricing

Vanna is free and open source (MIT); costs come from the LLM API you connect, and an optional hosted version is offered separately.

Frequently asked questions

What does Vanna do?

Vanna generates SQL from natural-language questions, using RAG trained on your database schema and past queries.

Is Vanna free?

Yes. Vanna is MIT-licensed open source; you pay only for the LLM API you choose to connect.

Which databases does Vanna support?

Vanna works with many common databases, including Postgres, MySQL, BigQuery, and Snowflake, via Python connectors.

Similar products

• Akkio — No-code platform for building predictive models and analyzing data with AI for business teams.
• Hex — Collaborative data workspace whose Magic AI helps write queries, build notebooks, and explore data.
• Julius — AI data analyst that answers questions, analyzes files, and creates charts from natural-language prompts.
• ThoughtSpot — Analytics platform with an AI assistant for natural-language business intelligence and search.

All Vanna alternatives → · All Data & Analytics products →

Sources

• Official website ↗
• Source code (GitHub) ↗

This record was last reviewed on 2026-07-30.

Machine-readable record: /api/products/vanna.json · Spot an error? Suggest a correction