DBQuill

DBQuill is a free, open-source Windows desktop app for asking database questions in natural language, reviewing operations, and viewing charts. It supports SQLite, along with read paths for MySQL 8.4 and PostgreSQL 17. Users can attach CSV or .xlsx files for conversion into a local SQLite database; legacy .xls files are not supported. Writes require validation, a change preview, and explicit confirmation, while remote DDL is blocked. Reads are limited to one statement and bounded rows, and SQLite reads use physical read-only connections plus query_only. DBQuill accepts an OpenAI-compatible text-model endpoint without requiring a particular model vendor. Credentials, model profiles, sessions, audit records, and uploads are stored locally outside the source tree. Audit records retain controlled metadata and hashes rather than raw prompts, SQL, credentials, or result rows. The current release target is Windows 10 or 11 x64 with CPython 3.12 and Microsoft Edge WebView2 Runtime. Distribution is source-first, and a signed native installer is not yet provided. macOS, Linux, Windows on ARM, and other Python versions are not release targets yet.

DBQuill pairs natural-language database queries with explicit write controls and local data storage. Its source-first distribution and limited Windows release target are important setup considerations.

It suits people working with SQLite, MySQL, or PostgreSQL data who want natural-language queries and controlled updates. It also supports tabular CSV and .xlsx files.

DBQuill is an open-source Windows desktop app for querying databases in natural language, reviewing changes, and viewing charts. It is best suited to people who want local data handling and a deliberate approval step before writes. Its guarded workflow is appealing, but source-first setup and narrow platform support make it a poor fit for users who need a signed installer or broader OS coverage.

Overview

DBQuill brings database questions, controlled updates, and result visualizations into a self-hosted desktop workflow. SQLite is supported, while MySQL 8.4 and PostgreSQL 17 have read paths; users should not expect remote write support from those connections. The app accepts OpenAI-compatible text-model endpoints rather than requiring a particular model vendor, which preserves choice but leaves endpoint setup to the user.

Credentials, model profiles, sessions, audit records, and uploads stay locally outside the source tree. Audit records retain controlled metadata and hashes, not raw prompts, SQL, credentials, or result rows. That is a meaningful privacy-conscious design for individual database work, though it is not a substitute for checking whether a deployment meets an organization's own controls.

Key features

Queries, files, and charts

Natural-language queries and result charts make DBQuill useful when a user wants to explore database content without composing every query manually. CSV and .xlsx attachments are converted into a local SQLite database, extending that workflow to tabular files without requiring a separate server. Legacy .xls files are unsupported, so older spreadsheet collections may need conversion first.

Writes with an approval gate

Writes require validation, a change preview, and explicit confirmation. This gives users a chance to inspect proposed changes before they take effect, a sensible safeguard for controlled updates. Remote DDL is blocked, and the MySQL and PostgreSQL paths are read-only in scope; DBQuill is therefore better suited to cautious exploration and supported local SQLite work than to broad remote schema administration.

Bounded reads

Reads are single-statement and row-bounded. SQLite reads also use physical read-only connections and query_only, adding database-level restrictions to the workflow. These controls narrow the ways a read can alter data, but users still need to choose appropriate database access and follow their own operational safeguards.

Pricing

DBQuill is free and open source: its DBQuill plan costs nothing. The plan includes natural-language queries, write operations, a semantic layer, result visualizations, and self-hosted deployment. There is no paid tier or free-trial decision to weigh, but the cost advantage comes with a source-first Windows setup rather than a signed native installer.

Platforms

The release target is Windows 10 or 11 x64 with CPython 3.12 and Microsoft Edge WebView2 Runtime. DBQuill does not yet target macOS, Linux, Windows on ARM, or other Python versions. Its source-first distribution and lack of a signed native installer make installation less straightforward than a packaged desktop app, so it is best for users comfortable following an installation guide and using diagnostics or GitHub issues for support. No response-time or security-report bounty commitment is offered.

Who it's for

DBQuill suits Windows users working with SQLite, or those who need read access to MySQL 8.4 or PostgreSQL 17, and who value local storage and explicit write review. It also fits people who want to bring CSV or .xlsx files into a local SQLite workflow. Look elsewhere if you need macOS or Linux support, remote database writes, a ready-to-install signed app, or a defined support response commitment.

Pros and cons

• Pros: Writes require validation, preview, and confirmation, reducing the chance of an unreviewed change.
• Pros: Credentials and work data stay local, while audit records avoid retaining raw prompts, SQL, credentials, or result rows.
• Pros: OpenAI-compatible endpoint support avoids dependence on one model vendor.
• Cons: Windows 10 or 11 x64 is the only current release target, excluding Mac, Linux, and Windows on ARM users.
• Cons: Source-first distribution without a signed native installer raises the setup burden.
• Cons: MySQL and PostgreSQL have read paths rather than a stated remote-write path, and remote DDL is blocked.

Alternatives

For a broader shortlist, see AI Database Assistants. Choose ThoughtSpot instead if web and mobile availability and paid plans starting at $25 a month billed annually better suit a 5-to-50-user deployment. DataZen is a free, GPLv3 alternative with Linux and macOS support as well as Windows, and it requires no account while letting users bring their own AI provider.

Florentine.ai is worth considering for a web or API workflow that includes a free tier capped at one database, five active collections or tables, and 500 requests per month. Insight O' Mate offers a free plan capped at 20 database analyses a day, one founder, and one database, with a $29 monthly Pro plan for unlimited queries, a priority queue, and advanced export. Those caps make it a clearer option for readers who want a defined entry-level allowance or paid upgrade.

Outerbase AI may suit users seeking web, API, macOS, Windows, or self-hosted access; its free plan caps usage at five users, one base, 10 EZQL queries a month, three saved queries, and one dashboard. Wren AI is an alternative for those interested in an open-source context engine for individual developers, used through CLI and MCP without a UI. YourQL is another free desktop application, described as a work in progress. DataLine is an open-source project with a public GitHub repository under GPL-3.0.

Verdict

Choose DBQuill if you use a supported Windows setup, want natural-language database work, and value local data handling plus an explicit checkpoint before writes. Its free open-source plan and vendor-flexible model endpoint are strong reasons to try it. Look elsewhere if you need wider platform coverage, a signed installer, remote write support, or a firmer support commitment.