Select Beta
SELECT SELECT · Tauri v2 + Rust · Beta

The lightweight, lightning-fast SQL client built for developers who care about speed.

Native Rust + Tauri v2 architecture. Zero bloat, under 60 MB RAM, instant launch, and zero Java/Electron lag. Query MySQL and MariaDB with sovereign local-first privacy.

macOS 12+ (Apple Silicon & Intel) · Windows 10 / 11 · Linux
query.sql
0.02ms
SELECT SELECT
FROM all_other_ides
WHERE simplicity = 100
  AND performance = 'instant'
  AND bloat = 0;
➜ Match: Select IDE
Select Desktop · courier_mesh STAGING RO
32MB RAM · 0.04s
Select Desktop SQL Editor

Engineered for developers querying high-throughput databases

MySQL 8.4
MariaDB 11
AWS RDS
PlanetScale
Docker
Cloud SQL

LIVING RUNBOOKS & NOTES

Turn ad-hoc queries into living runbooks. Notes & SQL unified in one editor.

Stop juggling Notion, Obsidian, and your database client. Write Markdown documentation, checklists (- [ ]), and incident runbooks alongside executable SQL blocks. Run queries in place with ⌘↵ without ever switching windows.

Markdown + Executable SQL Blocks Task Checklists (- [ ]) In-Place Query Run (⌘↵) Slash Commands (/) Living Database Runbooks
Context Switches
0 Windows
Format
Markdown + SQL
incident_runbook.note.sql · Notes & SQL
⌘↵ Executable
# Database Health & Replication Audit
Sprint 24 • Incident post-mortem • Primary EU-Central
☑ 1. Inspect connection pool saturation on replica nodes
☐ 2. Identify long-running uncommitted payment transactions (> 120s)
SQL BLOCK • ⌘↵ to execute 10 rows • 1.4ms
SELECT id, user_id, amount_cents, status, created_at
FROM orders
WHERE status = 'processing' AND created_at < NOW() - INTERVAL 2 HOUR
ORDER BY created_at ASC LIMIT 10;
Result: In-flight stuck orders detected in EU cluster. Webhook re-queued.
// Note: Runbooks are saved directly as `.sql` or `.note.sql` in your Git repo.
Select Editor Schema-Smart Autocomplete and Query Writing

SCHEMA-SMART INTELLIGENCE

Type SQL at the speed of thought. Schema-smart suggestions & visual EXPLAIN.

Real-time relational autocomplete as you type. Context-aware column data types, nullability badges, and auto-generated foreign key JOIN ... ON proposals inferred directly from your live schema. Inspect query performance with visual EXPLAIN plan cost graphs and format queries with ⌘⇧F.

FK-Aware JOIN Proposals Column Types & Nullability Table Alias Inference Visual EXPLAIN Cost Graph ⌘⇧F SQL Code Formatter
Suggestion Latency
< 1 ms
Schema Context
Zero Index Lag

60FPS VIRTUALIZED DATA GRID

Edit cells in place. Inspect foreign keys without losing your spot.

Scroll 100,000+ rows smoothly at 60 frames per second without DOM thrashing. Double-click any cell to edit with instant dirty-state tracking (visual amber diff bar), review generated SQL UPDATE scripts before committing, and hover over foreign keys to peek referenced rows in a floating popover. Safely update tables lacking primary keys using Virtual Unique Keys.

Amber Diff Tracking Bar FK Peek Popover Large Value JSON Drawer Virtual Unique Keys 1-Click TSV / JSON / SQL Export
Scroll Performance
60 FPS Smooth
Dataset Capacity
100k+ Rows Safe
Virtualized Data Grid with In-Place Cell Range Selection and Fast Copy
Schema ER Canvas · Relational Foreign Keys
Auto-Layout
Schema ER Canvas

RELATIONAL CLARITY

Relational schemas, mapped instantly.

Visualize foreign key dependencies, table relationships, and indexed schemas in real time. Pan, zoom, and isolate any table with targeted dependency focus mode. Inspect column constraints, indexes, and generated DDL without manual diagram setup or cloud subscriptions.

Real-Time FK Graph Canvas Targeted Table Focus Mode DDL & Index Deep Inspector Single-Query FK Resolution Zero-Config Ingestion

SOVEREIGN SECURITY & SAFETY

Triple-layer production guardrails and sovereign security.

Never accidentally drop a production database. Staging and production connections feature color-coded environment shields, triple-layer read-only locks, and a destructive query AST scanner that intercepts missing WHERE clauses before execution. Credentials are sealed in your native OS Keychain with AES-256-GCM, sensitive tokens in query history are masked automatically, and runaway queries can be killed instantly.

Native OS Keychain AES-256-GCM • Zero Cloud Telemetry
Triple-Layer Read-Only Shield Destructive Query AST Scanner OS Keychain AES-256-GCM Vault History Credential Redaction Thread-Safe KILL QUERY
Connection Manager · Safeguards
Read-Only Shield
Connection Manager and Safeguards
Select · Gruvbox Dark (Warm Retro + Film Grain)
⌥⌘>
Theme Preview
Switched to theme "Gruvbox Dark" (⌥⌘>)

LOCAL-FIRST WORKFLOW & 46 THEMES

Saved queries that live in Git, styled with 46 handcrafted themes.

Save queries as plain .sql files directly in any local directory on your drive, version-controlled with Git and shared with your engineering team. Cycle through 46 curated developer themes instantly using ⌥⌘> or the live switcher below. Featuring warm Gruvbox retro palettes with subtle organic film grain, CRT phosphor scanlines, and high-contrast daylight modes.

Git-Tracked Local .sql Directory 46 Curated Developer Themes Fuzzy Command Palette (⌘K) Tactile Audio Cues (cuelume) Pinned Result Snapshots (Up to 10)

Gruvbox Dark: Warm retro earthy palette (#282828) with subtle organic film grain, warm amber, moss green, and terracotta syntax.

The sovereign way to query databases.

Engineered from first principles without Java, Electron, or cloud telemetry.

“Consumes 95% less RAM than DBeaver and DataGrip. Keep Select running in the background with 30 open query tabs without feeling a single frame drop.”

Sub-32MB Memory -95% RAM Footprint

“Click the app icon and start typing SQL instantly. No JVM boot phases, no splash screens, and zero garbage collection UI lockups.”

Instant Cold Boot < 40ms Startup

“Write living runbooks with Markdown notes, task checklists, and executable SQL blocks in one unified tab. Run queries in place with ⌘↵.”

Living Runbooks Notes + SQL Hybrid

“60fps virtualized data grid renders 100k+ rows effortlessly. Edit cells in place with amber diff tracking and hover to peek referenced foreign keys.”

Virtualized Grid Amber Diff • FK Peek

“Interactive foreign key graph auto-generates from information_schema in real time. Click any table to isolate its relational ancestry in focus mode.”

Schema ER Canvas Table Focus Mode

“Saved queries live as plain .sql files in your Git repo. Triple-layer read-only shield, destructive AST checker, and native OS Keychain AES-256 vault.”

Sovereign Security Git Sync • Keychain
Dimension Select (Beta) DBeaver TablePlus DataGrip Beekeeper Studio MySQL Workbench
Runtime & Engine Tauri v2 + Tokio Rust Eclipse Java JVM Native C++ / Swift JetBrains JVM Electron / Node.js C++ / Python 3
Idle RAM Footprint < 32 MB 780 MB – 1.2 GB ~120 MB 650 MB – 1.5 GB 350 MB – 600 MB 540 MB
Cold Startup Latency < 40 ms (Instant) 9.40s ~350 ms 6.70s 3.20s 5.10s
Notes & SQL Runbooks Built-in Markdown + Executable SQL No No Scratch files only No No
Autocomplete & Suggestions Schema-Smart + FK JOINs Basic keyword / table Basic prefix matching Deep AST (Heavy index) Basic schema cache Simple keyword cache
Grid & In-Place Edits 60fps Virtualized + Amber Diff Standard (manual save) Inline cell editing Grid transaction commit Standard table view Clunky grid editor
Foreign Key Peek Popover Instant In-Grid Popover No Limited Navigation jump Navigation jump No
PK-Less Table Editing Virtual Unique Keys Safe Unsafe all-match Unsafe / blocked Unsafe / blocked Blocked Read-only
Large Value JSON Drawer Slide-out Formatter & Editor Modal dialog Cell popover Cell editor Basic viewer Basic viewer
Relational Schema ER Auto FK Graph + Focus Mode Static ERD (slow) None / basic view Heavy UML diagrammer None Manual modeler
Destructive Query Guard AST Scanner (Missing WHERE) Soft prompt Soft prompt Multi-row warning Soft prompt Safe update toggle
Production Read-Only 3-Layer Lock (UI+AST+Rust) Connection type (soft) Safe mode (basic) Read-only toggle Read-only toggle Read-only toggle
Saved Queries Storage Plain .sql in Local Git Repo Proprietary project Proprietary cloud/local IDE Project files Local SQLite / Cloud Project files
Query Cancellation Instant Thread KILL QUERY Often hangs socket Basic cancel IDE cancel Connection reset Buggy thread kill
Credential Storage OS Keychain (AES-256-GCM) Obfuscated master pass OS Keychain KeePass / OS Keychain OS Keychain OS Keychain
Cloud Telemetry 0% (Completely Local-First) Optional telemetry License phoning home JetBrains analytics Anonymous usage ping Oracle telemetry
License & Pricing MIT · Free Public Beta GPL / Paid Enterprise ($25+/mo) $89 per device lock $229 / year Open Source / $99 Comm. GPL (Oracle)

Download Select Beta

Available as native binaries for macOS, Windows, and Linux. Free & open-source under MIT.

Latest Beta · Release Notes →
macOS 12+

macOS

Universal DMG for Apple Silicon (M1/M2/M3/M4) and Intel Macs.

Gatekeeper Note (macOS):

Since beta builds are unsigned, right-click Select.app → Open, or run:

Alternative: Remove quarantine flag (all versions)

This bypasses Gatekeeper without going through System Settings:

xattr -cr /Applications/Select.app

After running this, double-click Select.app normally.

Win 10 / 11

Windows

Installer and standalone setup packages with WebView2 acceleration.

SmartScreen Note:

Click More Info → Run anyway. Microsoft WebView2 runtime required (standard on Windows 10/11).

Ubuntu / Fedora

Linux

Universal AppImage and Debian/Ubuntu packages for x86_64.

Make Executable:

chmod +x select_amd64.AppImage

Frequently Asked Questions

Everything you need to know about Select's architecture, security, and beta release.

Legacy editors run on the Java Virtual Machine (JVM) or wrap heavy Chromium/Electron runtimes that consume 600MB–1.2GB of memory just idling. Select is built with Tauri v2 and compiled to native Rust binaries. Database queries run asynchronously on the Tokio runtime with zero garbage collection sweeps.
Yes, 100% local-first. Select contains zero telemetry, zero analytics tracking, and zero cloud synchronization servers. Your database connection credentials, passwords, queries, and results stay entirely on your local machine.
Apple requires developers to pay an annual developer subscription to notarize binaries through Gatekeeper. Because Select is a free open-source project in public beta, the DMG is unsigned. You can open it by right-clicking Select.app and choosing Open, or running xattr -cr /Applications/Select.app in terminal.
Select is strictly optimized for MySQL (8.x & 5.7) and MariaDB (10.x & 11.x), along with hosted cloud variants like AWS RDS Aurora, Google Cloud SQL, PlanetScale, and TiDB. PostgreSQL and SQLite adapters are planned on our roadmap.
In Notes mode, your editor acts as a hybrid document. You can write Markdown headers, bullet points, checklists (- [ ]), and embed executable SQL code fences (```sql). Placing your cursor inside any SQL block and hitting ⌘↵ executes just that block and streams results below it, turning investigation notes into reproducible runbooks.
Yes! Unlike legacy tools that hide saved queries in proprietary binary blobs or cloud databases, Select lets you point your Saved Queries folder to any local Git repository. Queries are saved as plain .sql files, making them easily committable, branchable, and reviewable via pull requests.
When editing a table lacking a designated Primary Key, Select allows you to configure a Virtual Unique Key by selecting one or more identifying columns. When applying edits, updates are strictly parameterized and executed with LIMIT 1 to prevent unintended multi-row modifications.
When a connection is marked Read-Only, writes are intercepted at three distinct tiers: First, the UI grays out mutation actions. Second, a client-side SQL AST tokenizer parses incoming statements to block modifying verbs outside of comments/literals. Third, the native Tokio/Rust backend pool driver hard-rejects mutating queries before they ever hit the database wire.

Try Select now.

Join developers escaping bloated SQL editors. Free and local-first.