Wren AI

What is Wren AI?

Wren AI is a generative business intelligence platform for people and AI agents. It translates plain-language questions into SQL, executes queries against connected data, and presents summaries, result previews, charts, dashboards, and follow-up questions.

Its architecture separates the user interface, AI retrieval and SQL generation, and a semantic context layer called Wren AI Core. Business meaning is modeled explicitly through metadata, definitions, relationships, instructions, profiles, and remembered query examples. This helps the system generate queries from governed business context instead of exposing a raw schema directly to a language model.

How to Use Wren AI

1. Choose cloud or self-hosted deployment and connect a supported data source.
2. Model tables, fields, relationships, metrics, and business definitions in the semantic layer.
3. Add instructions and verified examples for ambiguous terminology.
4. Ask a business question in natural language and inspect the generated SQL before trusting the answer.
5. Validate results against known totals, filters, time zones, and access policies.
6. Save reliable queries as views, build charts or dashboards, and use follow-up questions for exploration.
7. For embedding, use the available SQL and chart APIs on a supported paid plan.

Text-to-SQL can produce valid but incorrect queries. Apply read-only credentials, row and column controls, query limits, cost guards, and human validation for financial, operational, or regulated decisions.

Core Features

• Natural-language analytics: Converts business questions into SQL and summarized answers.
• Governed semantic layer: Stores models, relationships, calculated fields, definitions, and instructions.
• SQL transparency: Shows generated SQL and lets users adjust it.
• Charts and dashboards: Generates visualizations and reusable BI outputs from query results.
• Follow-up exploration: Maintains conversational context for related questions.
• Broad data connectivity: Supports more than 20 sources, including major databases and warehouses.
• Embedding APIs: Provides SQL and chart generation for integration into other applications.
• Open-source core: Offers a free context layer, CLI, semantic engine, SDK integrations, and bring-your-own-LLM support.
• Deployment choice: Supports cloud and commercial self-hosted options.

Use Cases

• Let business teams answer routine data questions without hand-writing SQL.
• Create governed analytics copilots grounded in company definitions.
• Generate and iterate charts or dashboards from natural-language requests.
• Embed text-to-SQL and chart creation into an internal or customer-facing product.
• Give coding agents a semantic context layer for safer data access.
• Reuse confirmed natural-language-to-SQL examples across future questions.
• Support analytics across heterogeneous databases and warehouses.

Pricing

Wren AI offers an open-source option at no cost and paid cloud or commercial self-hosted plans. Current self-hosted Business pricing is listed from $899 per month when billed annually, based on concurrent sessions; Enterprise Plus uses custom pricing. Cloud plans use credits, and new accounts receive limited initial and monthly free credits. Verify current feature gates, credit consumption, hosting requirements, and API access before choosing a plan.

Frequently Asked Questions

Is Wren AI only for analysts?

No. Its natural-language interface targets business users, while SQL visibility, modeling, APIs, and self-hosting support analysts and developers.

Does Wren AI guarantee correct SQL?

No. Semantic context, validation, and query memory improve reliability, but teams must still review SQL and reconcile important results.

Can it be embedded in another product?

Yes. Paid plans provide APIs for natural-language SQL and chart generation. The API documentation currently notes that embedded API calls use Interactive Mode.

Is Wren AI open source?

Yes. The open-source offering includes the context layer, CLI, Rust semantic engine, framework SDKs, and bring-your-own-LLM support.