Natural Language SQL

Overview

Natural Language SQL turns questions written in plain English into database queries, then validates and runs them while returning results progressively. It is best suited to developers and teams who want a self-hosted, local-AI workflow with the option to inspect SQL before it runs. The open-source project offers useful privacy controls, but its Docker setup needs substantial disk space and write queries are permitted by default.

Key features

Hybrid schema retrieval combines BM25 search with sentence-transformer vectors to find relevant schema context for a question. That can save users from assembling database context manually, but generated SQL still needs scrutiny: the project reports 52.32% execution accuracy on Spider dev, not a guarantee that a particular query will be correct.

Users can review and edit a generated query before execution, and the interface streams progress through the workflow. Query explanations, sorting, 50-row pagination, CSV and JSON downloads, and tab-separated copying make results easier to inspect and pass along. SELECT queries receive a default LIMIT; explicit result counts are capped at 1,000 rows by default, and execution times out after 30 seconds by default.

These limits help bound queries, but they do not prevent changes to data. INSERT, UPDATE, and DELETE are allowed by default and accompanied by warning banners. To block writes rather than rely on users to avoid them, the project recommends connecting with a read-only database account.

Ollama is the default AI provider, keeping questions, schema, and generated SQL on the user's machine. OpenAI, Gemini, and Groq are optional cloud providers; choosing one sends the question, relevant schema portion, and SQL context to that provider, while database passwords and result rows are excluded. Passwords and cloud API keys are Fernet-encrypted at rest, redacted from logs, and masked in API responses. The application says it has no telemetry, and session query results, schema metadata, and history stay in memory rather than being written to disk by the application.

Pricing

Natural Language SQL Engine is free under the GPL-3.0 open-source license, for local installation. There are no paid tiers or published usage quotas to weigh against the free option; users take on the cost and work of running it themselves. The recommended Docker setup requires Docker 20.10 or later, Compose v2, and about 10 GB of free disk space, while the first run downloads an approximately 5 GB model. That footprint makes it a less convenient fit for machines with limited storage.

Platforms

Natural Language SQL supports API, Linux, macOS, self-hosted, and web use. It connects to PostgreSQL and MySQL, including multiple databases of the same type queried in parallel. That is useful for teams with those database systems, but it is not a fit for a workflow that depends on another database engine.

Who it's for

This is a strong fit for technically comfortable users who want to ask questions of PostgreSQL or MySQL, keep the default AI workflow local, and retain control over generated SQL. It is less suitable for teams seeking a managed service, a lightweight setup, or write protection without configuring database permissions. Cloud AI is an option for users willing to send query context to an external provider.

Pros and cons

• Pro: Local Ollama keeps questions, schema, and generated SQL on the user's machine, supporting a private default workflow.
• Pro: Editable SQL, query explanations, and streamed progress give users ways to inspect work before relying on results.
• Pro: PostgreSQL and MySQL support, including parallel queries across multiple same-type databases, suits users managing more than one database.
• Con: The recommended setup needs about 10 GB of free disk space, and its first run downloads an approximately 5 GB model.
• Con: Write queries are allowed by default, so preventing changes depends on using a read-only database account.
• Con: The reported Spider dev execution accuracy of 52.32% makes query review important rather than optional for consequential work.

Alternatives

Browse AI SQL Generators to compare tools across the category.

Choose Outerbase AI instead if you want a freemium option across API, macOS, self-hosted, web, and Windows, or a free plan capped at five users, one base, 10 EZQL queries per month, three saved queries, and one dashboard.

NatureQuery is another freemium option, with a web and API presence and a free plan that allows 50 monthly queries, one database connection, exports, and 30-day query history.

Consider QueryPlane if a self-hosted or web tool with a free trial and a free allowance of three apps, one database connection, and 50 AI prompts per month better matches your needs.

SQL Mocker offers a browser-based freemium alternative; its free Explore plan includes one user, 50 AI queries total, and five saved projects.

For a free open-source context engine that can be used through CLI and MCP without a UI, consider Wren AI.

mnemiq is a free, open-source engine that runs in your own environment.

Text2SQL.ai is a paid alternative with a free trial and API access that includes 100 requests per month on its Pro Plan.

For a freemium API, self-hosted, or web option, compare Vanna AI.

Verdict

Choose Natural Language SQL if you want a free, open-source way to query PostgreSQL or MySQL while keeping AI processing local and reviewing SQL before execution. Its main strengths are that privacy-oriented default and a transparent, editable workflow; look elsewhere if you need a lighter deployment or write queries blocked without managing database permissions.