Anyquery
Query anything (GitHub, Notion, +40 more) with SQL and let LLMs (ChatGPT, Claude) connect to using MCP
Last verified:
What is Anyquery?
Anyquery is an open-source SQL query engine that allows you to run SQL queries on pretty much anything – files, databases, APIs, and apps. It supports querying JSON, CSV, TSV, Parquet files, local/remote files (HTTP, S3, GCS), databases (SQLite, MySQL, PostgreSQL), and over 40 integrations including SaaS platforms like Notion, Google Sheets, Airtable, GitHub, Todoist, Apple Notes, Chrome tabs, Spotify, Salesforce, Shopify, Discord, Docker, and more. Built on top of SQLite, it uses plugins to extend functionality and can also act as a MySQL server for connecting BI tools like Metabase, Tableau, Looker Studio, and ORMs.
Key features include built-in MySQL server compatibility (runs on 127.0.0.1:8070 by default), JOIN support across APIs/files/SQLite tables, DML support (INSERT/UPDATE/DELETE) for Notion/Google Sheets/Airtable, log querying using Grok for parsing Apache/Nginx/Docker logs, alternative query languages (PRQL and PQL/Kusto-inspired), data visualization through MySQL compatibility with BI tools, query export to JSON/CSV/TSV/Markdown/HTML, and the Query Hub with pre-made SQL queries from the community. It also connects to LLMs (ChatGPT, Claude Desktop, Cursor, TypingMind) via Model Context Protocol (MCP) to let AI chat with your data.
Anyquery is designed for developers, data analysts, and anyone who loves SQL but wants to query data beyond traditional databases. It runs locally on your computer (data stays on your system), is free and open-source under AGPL-3.0 license, and can be installed via Homebrew, APT, YUM/DNF, Scoop, Winget, or Chocolatey. The tool is particularly useful for migrating data between SaaS platforms, exploring API data without learning complex API syntax, and leveraging the entire SQLite ecosystem including sqlite-vec, Datasette, and Litestream.
Anyquery pricing
Pricing model: Freemium
Anyquery is completely free and open-source. It is licensed under AGPL-3.0-or-later for the core engine (the RPC library is MIT licensed). There are no paid tiers, subscriptions, or premium plans. All features including the MySQL server, all 40+ integrations/plugins, Query Hub access, and LLM integration via MCP are available for free. The tool runs locally on your computer with no cloud fees or data transfer costs.
Anyquery pros
- Open source and completely free under AGPL-3.0 license
- Supports 40+ integrations including Notion, Google Sheets, Airtable, GitHub, Todoist
- Built-in MySQL server for connecting BI tools like Metabase and Tableau
- JOIN support across different data sources (APIs, files, databases)
- DML support (INSERT/UPDATE/DELETE) for Notion, Google Sheets, Airtable
- Query local and remote files: JSON, CSV, TSV, Parquet, HTML tables
- Log querying with Grok for Apache, Nginx, Docker logs
- Alternative query languages: PRQL and PQL (Kusto-inspired)
- LLM integration via MCP for ChatGPT, Claude, Cursor, TypingMind
- Query Hub with pre-made SQL queries from the community
- Export query results to JSON, CSV, TSV, Markdown, HTML
- Built on SQLite – compatible with entire SQLite ecosystem (sqlite-vec, Datasette, Litestream)
- Runs locally – your data stays on your system
- Easy installation via Homebrew, APT, YUM/DNF, Scoop, Winget, Chocolatey
- Plugin-based architecture – extendable with custom plugins
- Data visualization through MySQL compatibility with BI tools
- Transform Google Sheets/Airtable into SQL databases
Anyquery cons
- CLI-only interface – no native GUI application
- Requires SQL knowledge to write queries (unless using Query Hub)
- Some plugins may have limited functionality compared to native APIs
- DML support currently limited to Notion, Google Sheets, Airtable only
- Active development – some features may be in beta (e.g., database joins from v0.3.0)
- Only one main developer – contributions currently limited to minor ones
- AGPL-3.0 license may be restrictive for commercial proprietary use
- No cloud/saas version – must run locally on your machine
- LLM integration requires separate MCP client setup
Frequently asked questions about Anyquery
What is Anyquery?
Anyquery is a SQL query engine that allows you to run SQL queries on pretty much anything – files (JSON, CSV, Parquet), databases, APIs, and apps (Notion, Google Sheets, Airtable, GitHub, Todoist, Apple Notes, Chrome tabs, Spotify, etc.). It's built on top of SQLite and uses plugins to extend its functionality. It can also connect to LLMs via MCP and act as a MySQL server for BI tools.
Is Anyquery free?
Yes, Anyquery is completely free and open-source. It is licensed under AGPL-3.0-or-later for the core engine. There are no paid plans, subscriptions, or premium features. All integrations, the MySQL server, Query Hub, and LLM integration are included for free.
How do I install Anyquery?
You can install Anyquery via Homebrew (brew install anyquery), APT, YUM/DNF, Scoop, Winget (winget install JulienCagniart.anyquery), or Chocolatey (choco install anyquery). You can also download binaries from the GitHub releases page. The documentation provides detailed installation instructions for each platform.
What data sources does Anyquery support?
Anyquery supports 40+ integrations including: files (JSON, CSV, TSV, Parquet), databases (SQLite, MySQL, PostgreSQL), SaaS apps (Notion, Google Sheets, Airtable, GitHub, Todoist, Salesforce, Shopify, Spotify, Discord, ClickUp, Trello, Hubspot), local apps (Apple Notes, Apple Reminders, Chrome/Firefox/Edge tabs, Docker, Git), logs (Apache, Nginx), and more. See the integrations page for the full list.
How do I connect Anyquery to LLMs like ChatGPT or Claude?
Anyquery connects to LLMs via the Model Context Protocol (MCP). Start the MCP server with 'anyquery mcp --stdio' or 'anyquery mcp --host 127.0.0.1 --port 8070'. For function calling clients like ChatGPT or TypingMind, run 'anyquery gpt' to get an ID to paste in the LLM client. Supported LLMs include ChatGPT, Claude Desktop, LibreChat, TypingMind, and 5ire.
Can I use Anyquery without knowing SQL?
Yes, you can use the Query Hub (anyquery.dev/queries) which contains pre-made SQL queries from the community. You can run queries using 'anyquery run' command with a query from the hub. The tool will prompt you for necessary information. However, for custom queries, SQL knowledge is required.
How do I use the MySQL server feature?
Launch the MySQL server with 'anyquery server' command. It starts on 127.0.0.1:8070 without authentication by default. Connect with any MySQL client like 'mysql -u root -h 127.0.0.1 -P 8070'. You can configure host/port with --host and --port flags, add authentication with --auth-file, and open specific databases with --database flag.
Does Anyquery support INSERT/UPDATE/DELETE operations?
Yes, Anyquery supports DML (INSERT/UPDATE/DELETE) for Notion, Google Sheets, and Airtable. You can use these platforms as SQL databases with full read/write capabilities. This feature allows you to modify data in these SaaS platforms using standard SQL syntax.
What query languages besides SQL are supported?
Anyquery supports alternative query languages: PRQL (Pipe Relational Query) and PQL (a Kusto-inspired language). If you don't like SQL, you can use PRQL. If you're used to Microsoft Kusto, you can use PQL. All query languages leverage the same underlying SQLite engine.
How do I export query results?
Anyquery allows you to export query results to multiple formats: JSON, CSV, TSV, Markdown, and HTML. You can use the query export feature to save your results in your preferred format for further analysis or sharing.