Tables 0.3.0 · macOS, Linux & Windows beta

Your data.
Room to work.

Browse tables, write SQL, and review changes in a compact native workspace. Keep each table’s filters and edits close, with optional assistant and read-only MCP support.

Download Tables →Read the docs

PostgreSQL
MySQL / MariaDB
SQLite

SQL editorcustomersorders ×

RefreshInsertFilterExport

DataStructure

id	customer	status	total
48291	Amara Okafor	processing	184.20
48288	Marc Vidal	review	942.00
48279	Yuna Park	processing	218.45
48264	Nora Singh	shipped	128.10
48251	Leo Anders	review	516.80

PostgreSQL48,902 rows · Page 1

Workspace illustration · sample data

Unified table tabs

Postgres · MySQL · SQLite

Zero telemetry

Read-only MCP

Built for the loop

See it. Shape it.
Ship it with confidence.

Browse

Move through real data without the drag.

Server-side paging, sorting, filters, and a dense native grid keep large tables legible.

Query

Write SQL where the answers land.

Run multiple statements, revisit history, save favorites, and turn results into quick charts.

Change

Review the SQL before it becomes reality.

Stage cell edits and deletes, inspect the generated statements, then commit the batch.

⌁

A grid with a memory.

Keep several tables open, each with its own page, filters, selection, and staged edits. Switch Data/Structure in the toolbar below. Review pending changes before committing them.

Read the data guide →

Review changes3 pending

UPDATEordersstatus = 'shipped'where id = 48288

UPDATEorderstrackingcode = 'TRK-88419'where id = 48288

DELETEorderswhere id = 48107

customercohorts.sql⌘ ↵ Run

select date_trunc('month', createdat) as cohort,
       count(*) as customers,
       round(avg(lifetimevalue), 2) as avgvalue
from customers
where createdat > now() - interval '1 year'
group by 1 order by 1;

SQL stays close to the data.

A focused editor, multi-statement results, history, favorites, and exploratory charts. No browser runtime between your keystroke and the query.

Read the query guide →

One workspace.
Different engines.

Tables keeps the workflow consistent while respecting the systems underneath.

PostgreSQL

Native types, schemas, indexes, foreign keys, and SSL.

MySQL + MariaDB

One browsing and editing flow across the protocol family.

SQLite

Open a local file and work instantly.

Your tools.
Your database.

The optional assistant helps you work with SQL and schema context. The standalone MCP server lets a client inspect saved connections, schemas, and row pages through read-only tools.

No running desktop app required for MCP. No HTTP listener. Explicit limits on requests, rows, and responses.

Set up the MCP server →

Four focused toolsstdio · read-only

list_connections
list_tables
table_schema
table_rows

New in 0.3.0

Unified titlebar tabs. Light and dark appearance. Direct toolbar actions. Bounded queries and assistant streams. Streaming table exports. Atomic metadata writes.

Read the release notes →How updates work →

Keep the database
within reach.

Fast enough for the daily loop. Transparent enough for the important changes.