Skip to content
RED

Roughly Enough Data

RED is a lightweight database explorer for developers who prefer fast, focused tools.

Try REDGView on GitHubS

Open source · Available for macOS, Linux and Windows

Native

Built with GPUI for a fast, responsive desktop experience.

Streaming

Results stream directly into a virtualized grid. Browse millions of rows without blowing through memory.

Safe

Read-only connections, query timeouts, and confirmation before destructive statements. Safer production inspection.

A familiar layout.

Browse the schema, write SQL, inspect results. The same window browses Redis keys and MongoDB documents.

RED's connected workspace - schema sidebar, SQL editor, a virtualized result grid streaming millions of rows, and the read-only AI agent. The same window browses Redis keys and MongoDB documents.RED's connected workspace - schema sidebar, SQL editor, a virtualized result grid streaming millions of rows, and the read-only AI agent. The same window browses Redis keys and MongoDB documents.

Open on a wider screen to try the live, interactive demo.

Rows, keys, documents.

RED speaks all three natively. Same grid, same shortcuts, same assistant, whatever shape your data is in - no second app, no second mental model.

SQL

Rows and relations

Browse the schema, write SQL with completion and inline diagnostics, and scroll a result that pages in as you go. Follow a foreign key to the row it points at, read the whole schema as an ER map, or diff it against another database.

  • PostgreSQL
  • MySQL
  • MariaDB
  • SQLite
  • ClickHouse
REDIS

Keys and streams

Scan a keyspace by pattern, prefix or value, edit a hash field in place, and read streams alongside their consumer groups. RedisJSON documents open a level at a time, so a 200 MB value costs no more than a small one.

  • Strings
  • Hashes
  • Sets
  • Sorted sets
  • Streams
  • RedisJSON
MONGODB

Documents and collections

Read a collection as a table, a card list or raw JSON, edit a document as a form or as text, and page through millions on an _id cursor. Indexes, validators and explain plans sit one click away.

  • Find
  • Aggregate
  • Explain
  • Indexes
  • Validators

Roughly enough features.

What a developer actually reaches for when inspecting a database - and a deliberate absence of everything else.

View roadmap
BROWSE

Millions of rows, flat memory

Results stream into a virtualized grid over a windowed cursor, so a 50-million-row table opens as fast as a small one. Size, hide and reorder columns, select a range, and copy it as TSV, CSV, JSON, Markdown, an IN (…) list or INSERT statements.

EXPLORE

Read the whole database

Tables, views, indexes, foreign keys, routines, triggers and types in one lazily loaded tree. Show any object's DDL, map the schema as an ER diagram, diff it against another, or run a health report for unused indexes and missing keys.

ASK

An agent that shows its work

⌘L asks Claude about your schema and data, on SQL, Redis or MongoDB. Every answer cites the queries behind it, every write waits for approval with the row count it would touch, and it stays dormant until you sign in.

PROTECT

Safe by default, editable on purpose

Connections are read-only until you say otherwise. Edits stage into a change-set you review and commit as one transaction, destructive statements ask first, and passwords live in your system keychain, never a plaintext config.

MOVE

Move data between anything

Import CSV and JSON, copy a table across connections, or migrate a whole schema with its keys and indexes. Export streams row by row to CSV, JSON, SQL INSERTs or Excel, and never holds the result in memory.

ALSO IN THE BOX⌘K command palette⌘P connection switcherSplit panesSession restoreQuery historySaved queriesWatch modeColumn statisticsServer & session monitorSSH tunnelsKeymap editorLive settings.tomlThemesCommand-line modeMCP serverNo telemetry

Free and open source.

No accounts. No telemetry. No lock-in. Just a database explorer you can inspect, modify, and build yourself.

GPL-3.0-or-later
View on GitHubS
red - zsh
$ git clone https://github.com/vojir-mikulas/red
Cloning into 'red'... done.
$ cargo run -p red
Compiling red v0.1.0
Finished `release` profile [optimized]
▸ RED ready - Roughly Enough Data
new-user@red $ 

Frequently asked questions

Read the docs

A native desktop app for working with databases. Connect to a server, browse the schema, write SQL, and scroll through results that page in as you go, so large tables open as fast as small ones. It's built in Rust on GPUI.