SQL Chat

English | 中文 | Español

SQL Chat is a chat-based SQL client, which uses natural language to communicate with the database to implement operations such as query, modification, addition, and deletion of the database.

As we enter the Developer Tools 2.0 era, there is a massive opportunity to rebuild the existing tools using the chat-based interface. SQL Client is no exception. Instead of navigating across many UI controls, a chat-based interface is much more intuitive. Of course, only if that works, and our goal is to deliver that experience.

SQL Chat is built by Next.js, it supports the following databases and will add more over time:

• MySQL
• PostgreSQL
• MSSQL
• TiDB Cloud

If you use sqlchat.ai to connect to your database, you need to add 0.0.0.0 (allow all connections) to the database whitelist IP. Because sqlchat.AI is hosted on Vercel which uses dynamic IP. If this is a concern, please consider the self-host option below.

See SQL Chat Privacy Policy.

If you just want to run for your own use, supply the following options:

• NEXTAUTH_SECRET
• OPENAI_API_KEY

docker run --name sqlchat --platform linux/amd64 --env NEXTAUTH_SECRET="$(openssl rand -hex 5)" --env OPENAI_API_KEY=<<YOUR OPENAI KEY>> -p 3000:3000 --hostname localhost sqlchat/sqlchat

• Pass an arbitrary string to NEXTAUTH_SECRET otherwise next-auth will complain.
• If you chat to the database on the same host, you need to use host.docker.internal as the host in the database connection setting.

• If you just want to use for yourself, then run without database. Check .env.nodb.
• If you want to offer a similar service as sqlchat.ai, then run with database, check .env.usedb. The database is used to store account, usage info.

• OPENAI_API_KEY: OpenAI API key. You can get one from here.

• OPENAI_API_ENDPOINT: OpenAI API endpoint. Defaults to https://api.openai.com. Use Ollama to set up self-host AI model and set the endpoint to it.

• NEXT_PUBLIC_ALLOW_SELF_OPENAI_KEY: Set to true to allow users to bring their own OpenAI API key.

• NEXT_PUBLIC_USE_DATABASE: Set to true to start SQL Chat with database. This will enable following features:
  1. Account system.
  2. Per-user quota enforcement.
  3. Payment.
  4. Usage data collection.
• DATABASE_URL: Applicable if NEXT_PUBLIC_USE_DATABASE is true. Postgres connection string to store data. e.g. postgresql:https://postgres:YOUR_PASSWORD@localhost:5432/sqlchat?schema=sqlchat.

1. Install dependencies

   pnpm i

2. Generate prisma client

   pnpm prisma generate

3. Make a copy of the example environment variables file:

   cp .env.usedb .env

4. Add your API key and OpenAI API Endpoint(optional) to the newly created .env file.

You can skip this section with NEXT_PUBLIC_USE_DATABASE=false if you don't build features requiring database

1. Start a Postgres instance. For mac, you can use StackbBricks, DBngin or Postgres.app.

2. Create a database:

   CREATE DATABASE sqlchat;

   In .env file, assign the connection string to environment variable DATABASE_URL and DATABASE_DIRECT_URL. This article explains why we need two URLs.

3. Set up database schema

   pnpm prisma migrate dev

4. (Optional) Seed data

   pnpm prisma db seed

This project is under the BSL License. See the LICENSE file for the full license text.

ping api.openai.com