Conar is a desktop application that aims to combine local-first database work with cloud-based AI assistance. It stands out by integrating AI models directly into a SQL IDE environment, supporting multiple databases, and encrypting connection data stored in the cloud. This makes it a practical tool for developers who want AI-powered query help without sacrificing security or the responsiveness of a local application.
What conar does and how it is built
Conar is an open-source project built primarily with TypeScript, React, and TailwindCSS for the frontend, packaged as an Electron desktop app. It supports native connections to several popular database systems: PostgreSQL, MySQL, Microsoft SQL Server (MSSQL), and Clickhouse. Users can manage their database connections securely as connection strings are encrypted and stored in the cloud.
The AI assistance is powered by several large language models from different providers including Anthropic, OpenAI, Gemini, and XAI. This model-switching capability allows users to choose or experiment with various AI-powered SQL generation and query optimization tools within the same interface.
Under the hood, the backend is implemented with Bun, a fast JavaScript runtime. It uses Hono as the HTTP framework, oRPC for remote procedure calls, and Drizzle ORM for database interactions. The entire project is organized as a Turborepo monorepo, which helps manage the multiple frontend and backend packages cohesively.
The architecture reflects a hybrid design: the desktop app runs locally for quick interactions and UI responsiveness, while cloud storage secures connection data and AI models provide the heavy lifting for SQL assistance. This approach balances local speed and offline usability with cloud convenience and AI power.
What sets conar apart technically and its tradeoffs
Conar’s key technical strength lies in its AI-native SQL IDE concept. Unlike traditional database tools that focus purely on schema browsing and manual SQL editing, Conar integrates multiple AI models directly into the workflow. This lets developers generate, refactor, and optimize SQL queries with natural language prompts or assisted editing. Such functionality is still emerging in the database tooling space.
Supporting multiple AI providers is both a strength and a complexity. It allows users to pick the best model for their needs or cost constraints, but it requires maintaining integrations with different APIs and managing diverse response formats. This flexibility reflects a practical tradeoff: more maintenance work in exchange for broader AI access.
From a database support perspective, Conar targets four distinct systems, which cover broad use cases from transactional databases (PostgreSQL, MySQL, MSSQL) to analytical workloads (Clickhouse). This multi-database support increases the user base but also complicates the UI and backend logic since each system has unique connection parameters and SQL dialects.
The use of an Electron desktop app combined with cloud-stored encrypted connection strings is a design choice that balances local-first performance with centralized management. The tradeoff includes the complexity of synchronizing secure data between client and cloud and the inherent overhead of Electron apps. However, Electron’s cross-platform nature and React’s UI flexibility make this a sensible choice for wide adoption.
The backend stack—Bun with Hono, oRPC, and Drizzle ORM—is a modern and performant JavaScript/TypeScript ecosystem. Bun’s speed benefits are appealing in development and may reduce resource consumption compared to Node.js. The use of Turborepo helps in managing monorepo complexity, improving developer experience.
Code quality appears solid given the choice of TypeScript throughout, which enforces type safety in both frontend and backend. TailwindCSS provides utility-first styling, keeping the UI code maintainable. The monorepo structure suggests a mature development process with clear separation of concerns.
Development setup and quickstart
Getting started with Conar’s development environment involves a few straightforward steps. The README provides exact commands to install dependencies, start supporting services, run database migrations, and launch the development servers:
pnpm install
pnpm run docker:start
pnpm run drizzle:migrate
pnpm run dev
pnpm installfetches all dependencies. Conar uses pnpm, which is faster and more disk-efficient than npm or yarn.pnpm run docker:startuses Docker Compose to launch PostgreSQL and Redis containers needed for development.pnpm run drizzle:migrateapplies database migrations using Drizzle ORM to prepare the schema.pnpm run devlaunches all development servers with Turbo, starting both backend and frontend concurrently.
This setup confirms the use of PostgreSQL as a development database alongside Redis, likely for caching or session management. The use of Docker Compose abstracts away manual setup of these dependencies.
verdict
Conar is a practical and modern AI-assisted SQL IDE that caters to developers who juggle multiple database systems and want AI help generating and optimizing queries. Its hybrid local/cloud architecture is well thought out, offering a responsive desktop experience without sacrificing secure cloud storage of connection info.
The multi-model AI integration is a notable feature, giving users flexibility but also introducing complexity that may affect stability and maintenance.
As an Electron app, it carries the typical tradeoffs of resource use and app size but benefits from cross-platform reach and a rich React UI.
This project is relevant for database developers and analysts who want to experiment with AI-assisted query building and work across PostgreSQL, MySQL, MSSQL, and Clickhouse. It’s also a useful case study in combining modern JavaScript runtimes like Bun with AI APIs in a desktop context.
If you’re looking for a slick AI-native SQL IDE that supports multiple databases and you don’t mind setting up a development environment with Docker, Conar is worth exploring. For production use, be aware of the usual Electron app limitations and the maintenance overhead of supporting multiple AI providers.
Overall, Conar is a solid, developer-focused tool that blends local-first design with cloud convenience and AI power in a pragmatic way.
Related Articles
- OrcaQ: a family-aware multi-database editor with safe AI assistance — OrcaQ is a Vue 3 and Tauri-based desktop database editor that adapts its UI for SQL and Redis. It offers schema explorat
- Foxel: a pluggable, AI-powered self-hosted cloud storage platform — Foxel offers a self-hosted cloud storage platform with runtime plugin loading and AI-driven semantic search across multi
- Supabase: composable open-source backend-as-a-service built around Postgres — Supabase combines specialized open-source tools around Postgres to offer a Firebase-like backend platform. Its modular a
- QA-Use: AI-powered natural language E2E testing platform with autonomous browser agents — QA-Use enables natural-language E2E tests using AI agents that autonomously interact with web apps. Built with TypeScrip
- Open Deep Research: A Next.js 16 agentic AI assistant for iterative web research — Open Deep Research is a TypeScript Next.js 16 app that uses an LLM to plan, execute, and iterate web research via Exa an
→ GitHub Repo: wannabespace/conar ⭐ 1,382 · TypeScript