Quary – Open-Source BI for Engineers

TL;DR

TL;DR: Quary is an open-source, SQL-first business intelligence platform built for engineers — connect your database, write SQL models, and build charts directly from VSCode.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: quary.dev
• Source repository: github.com/quarylabs/quary
• License: Apache-2.0 (verified via GitHub API)
• HN launch thread: news.ycombinator.com/item?id=40367090
• Source last checked: 2026-07-18

What Is Quary?

Quary describes itself as “Business Intelligence for Engineers.” Unlike traditional BI tools that hide SQL behind drag-and-drop UIs, Quary puts SQL at the center of the workflow. You connect a database, write SQL queries to model your data, and Quary renders charts and dashboards from those models — all within VSCode.

The core workflow:

1. Connect to your database (PostgreSQL, BigQuery, Snowflake, Redshift, Supabase, DuckDB, SQLite)
2. Write SQL to transform raw tables into analysis-ready models
3. Create charts directly from SQL results
4. Deploy models back to the database with version control

Quary was part of Y Combinator’s Winter 2024 batch.

Setup Workflow

Prerequisites

• VSCode
• A supported database (PostgreSQL, BigQuery, Snowflake, Redshift, Supabase, DuckDB, or SQLite)

Step 1: Install the CLI

macOS / Linux via curl:

curl -fsSL https://raw.githubusercontent.com/quarylabs/quary/main/install.sh | bash

macOS via Homebrew:

brew install quarylabs/quary/quary

Or download a binary from the releases page.

Step 2: Install the VSCode Extension

Install the Quary VSCode Extension from the VSCode Marketplace. The extension depends on the CLI being installed first.

Step 3: Initialize a Project

qr init

This creates a quary/ directory in your project with configuration files for your database connection.

Step 4: Connect Your Database

Edit quary/project.yaml to add your database connection. For example, connecting to PostgreSQL:

connections:
  mydb:
    type: postgres
    connection_string: postgresql://user:password@localhost:5432/mydb

Step 5: Write Your First Model

Create a SQL file in quary/models/:

-- quary/models/active_users.sql
SELECT
  user_id,
  created_at,
  last_login
FROM users
WHERE status = 'active';

Step 6: Create a Chart

Charts are defined in .sql.chart.yaml files alongside your models:

type: line
model: active_users
x: created_at
y: user_id
title: Active Users Over Time

Supported Databases

Quary supports the following databases out of the box:

| Database | Badge | |---|---| | PostgreSQL | postgresql:// | | Google BigQuery | bigquery:// | | Amazon Redshift | redshift:// | | Snowflake | snowflake:// | | Supabase | supabase:// | | DuckDB | duckdb:// | | SQLite | sqlite:// |

Deeper Analysis

Quary’s SQL-first approach is intentional. Traditional BI tools like Tableau or Looker abstract SQL away, which works for business analysts but frustrates engineers who need precision and reproducibility. Quary embraces SQL as the DSL for data modeling.

The asset types Quary manages as code are:

• Sources — external data inputs (tables, flat files, APIs via DuckDB)
• Models — SQL transformations that build on sources
• Charts — visual representations of model results

Dashboards and Reports are listed as work-in-progress on the README.

The tool is Rust-based under the hood (the CLI), with a VSCode extension providing the UI. This means the CLI can run headless in CI/CD pipelines, making it possible to validate models in pull requests.

Practical Evaluation Checklist

• [ ] Connects to a local PostgreSQL database
• [ ] Runs a SQL model through qr run
• [ ] Renders a chart from a model result in VSCode
• [ ] Uses version control on quary/ project directory
• [ ] CI/CD pipeline validates models on pull request

Security Notes

• Connection strings for databases are stored in project.yaml — never commit this file to version control if it contains credentials. Add project.yaml to .gitignore and use environment variables or a secrets manager for credentials.
• Quary does not appear to have a hosted/cloud offering based on the README; everything runs locally.
• The project is Apache-2.0 licensed with no commercial restrictions mentioned.

FAQ

Q: Does Quary support cloud databases like Snowflake and BigQuery? A: Yes. The README explicitly lists Snowflake, Google BigQuery, Amazon Redshift, and Supabase as supported databases.

Q: Can I use Quary without VSCode? A: The CLI works headless, so you can run qr run and other commands from the terminal. The VSCode extension provides the chart visualization UI. If you only need the CLI, you can use it without the extension.

Q: Is there a hosted or managed version? A: No. Quary is self-hosted and runs entirely on your local machine or server. There is no SaaS offering mentioned in the README.

Q: How does Quary compare to tools like dbt? A: Both are SQL-first data modeling tools. dbt is primarily a transformation layer that runs in data warehouses. Quary adds a visualization layer (charts, dashboards) on top of SQL models and integrates directly with VSCode. They can be complementary.

Conclusion

Quary is a practical open-source BI tool for engineers who think in SQL. It brings version-controlled data modeling and basic charting into the VSCode workflow, targeting teams that want BI-style insights without abandoning their SQL queries. If you already live in dbt or Looker it may overlap, but for teams wanting a lightweight, self-hosted BI layer with a minimal footprint, Quary is worth evaluating.

Links:

• quary.dev
• GitHub – quarylabs/quary
• VSCode Extension