Query files directly.
Explore local CSV, Parquet, and JSON files with SQL, right inside DataWorx.
BETA · BUILD 454
The beta is free to try.
New: DataWorx keeps itself up to date. From build 441 on, DataWorx checks for a new build when it starts and every few hours. When one is out, it shows you what’s new, and Install and Restart puts it in place. Help › Check for Updates… checks any time. If you have an earlier build, download this one once; new builds come to you after that. Also new: make an edited query a table’s default, so opening that table always starts with your query.
Recently added: Amazon Athena, Firebird and SurrealDB. Connect to an Athena workgroup with an AWS profile or access keys, find workgroups with Add from Cloud, and see what each query scanned and about what it cost. Connect to Firebird 3, 4 or 5 with nothing to install, run PSQL with or without SET TERM, and read the profiler’s measured plans. Connect to SurrealDB over WebSocket or HTTP, run SurrealQL with each statement’s rows as a table, and edit records by their IDs. Each comes with Explain, Server Activity, import, copy, design, dump, compare and search.
Also new recently: disk space for Amazon RDS, Aurora and DocumentDB in Database › Storage; Trino, Presto and SAP HANA; Compare for every database; Users and Roles, and Dump and Restore, for nearly every database; editing in nearly every database; Microsoft Entra ID sign-in and Microsoft Fabric. See every database and what it supports ↗
DataWorx-0.1.0-454.dmg · 96 MB · macOS 11 or later on an Apple silicon Mac (M1 or newer). This build doesn't run on Intel Macs. Experimental Windows builds are below.
Windows: Experimental Windows builds are below, in Windows & Linux testing. A Linux build may follow.
Serious database work. One complete workspace.
Write SQL, design schemas, debug routines, chart results, and work with AI.
FAAAASSSST by design. DataWorx opens instantly and stays responsive with large results, so your database workflow never waits.
Built by a software developer with four decades of database and software development experience.
AT HOME WITH YOUR DATABASES
INTEGRATED NOW / DUCKDB
DuckDB is already integrated into DataWorx. Its embedded analytics engine turns your database workbench into a place to work with data wherever it lives.
Explore local CSV, Parquet, and JSON files with SQL, right inside DataWorx.
Join PostgreSQL results with Snowflake results or a file on your machine through the integrated DuckDB engine.
Work with files and database results in the same workspace, with DuckDB built in and ready for analysis.
NEW NOW / MONGODB
MongoDB is now part of DataWorx: write MQL just as you would in the MongoDB shell, with an editor that knows your collections and fields and helps you build every stage of an aggregation.
find, aggregate, inserts, updates and indexes, exactly as the shell takes them. Autocomplete suggests collections, the fields in your documents, and the $ operators that fit where you are.
Choose a stage and get its whole phrase with placeholders to Tab through. See every form of the operator you're writing. After a $group or $lookup, the fields offered are the ones that exist at that point.
Results open as MongoDB JSON, or as a table with a column per field. Edit sub-documents as JSON, without drilling down; types are kept and conflicting changes are never overwritten.
DATABASE CONNECTIONS
Relational, analytics, NoSQL, graph and time series databases, each with a connection of its own: its own catalog, types, plans and sign-in. Databases that speak their protocols work too, and more are on the way. See every database ↗
01 / RELATIONAL
Purpose-built connections, with each database’s own dialect, catalog and plans, and its own sign-in and connection options.
02 / ANALYTICS & CLOUD
Dedicated connections for interactive analytics and cloud-scale queries, with each warehouse’s own plans in Explain.
03 / A DEDICATED NOSQL TRACK
Document and collection explorers, native query consoles, and native data display panels designed around each store’s own query language and data model.
WINDOWS & LINUX / IN TESTING
Windows and Linux versions are in testing, and we’re actively seeking testers. Put DataWorx through your everyday database workflows and help us refine the desktop experience on both platforms.
Experimental Windows builds. These are untested: they haven’t been run on Windows yet, so expect rough edges, and please tell us what breaks. They aren’t signed yet either, so Windows will warn you before installing. Choose More info, then Run anyway.
Experimental and untested · Build 454 · x86 (53 MB) for most Windows 10 and 11 PCs with Intel or AMD chips · Arm (47 MB) for Windows 11 on Snapdragon PCs like the Surface Pro 11. ODBC connections aren’t available on Arm yet. From this build on, Windows builds update themselves too.
Test real queries, connections, and workflows on the platform you use.
Become a tester ↗BUILT FOR THE WORK YOU DO
A complete workflow, from connecting to a database to sharing the answer.
Contextual completion, dialect-aware highlighting, formatting, and name inspections keep your thoughts moving. Save reusable queries with tags and Git-backed folders.
Explore results in a flexible grid. Sort, search, transpose, and inspect JSON or XML. Stage edits, review them, then submit the batch with transactional conflict detection.
Design tables visually, preview the SQL, and compare table structures across connections. Explore foreign-key relationships in interactive ER diagrams.
Chart query results as bars, lines, scatter plots, or pies. Inspect execution plans and actual query timings to understand how your SQL runs.
AI is available, but DataWorx will never force it on you. Use it when it helps, or work entirely with the database tools you know. Chat with your choice of AI engine using SQL and schema context. Connect external AI clients through MCP, with explicit connection access and read-only defaults.
Import CSV, JSON, Excel and Parquet, dump and restore with database-native tools, generate realistic test data, and export to Excel, JSON, SQL, Markdown, HTML, XML, or custom delimited formats.
Watch every session live with blocking chains drawn as a tree, cancel or kill with the SQL shown first, manage users, roles and permissions, and see table and index sizes with unused indexes flagged.
Dry Run any change and see exactly which rows it would touch. Production guards catch a missing WHERE, manual transactions wait for your Commit, and grid edits land all together or not at all.
Search every table, column and routine by name, or search inside the code of views, functions and triggers across a whole server. Every query you have run is a keystroke away in History.
System keychain credentials, SSH tunnels, TLS settings, and connection colors that help you know exactly where you’re working.
THE FULL PICTURE
339 documented capabilities across 15 areas. Explore the complete DataWorx feature catalogue.
Browse all 339 capabilities below. Use your browser’s Find command to locate a feature.
MySQL protocol adapter
Direct client, with no Instant Client to install; service names, SIDs and tnsnames.ora aliases.
TDS client
Self-hosted or CockroachDB Cloud, with its own plans in Explain and its system databases kept out of the way.
Serverless workgroups and provisioned clusters; distribution and sort keys in table details, and Redshift’s plans in Explain.
Self-hosted or ClickHouse Cloud over HTTP or HTTPS; results in ClickHouse’s own types, and Cancel stops the query on the server.
A project’s datasets, tables, views and routines, signing in with your Google account or a service account key; whole scripts run as one job.
See what a query would process before it runs and what each one billed after, and set a maximum bytes billed per connection.
Sources listed apart from tables, definitions with their clusters, its plans in Explain, and storage and indexes in Table Statistics.
Tables, sources and materialized views from its own catalog; rows you write are there for your next query.
Each table’s key model, distribution and partitioning; their plans in Explain; StarRocks primary-key tables editable in the grid.
Keys, checks and sharding in table details, shard sizes in Table Statistics, and rows you write there at once.
SQL warehouses with Unity Catalog’s catalogs, schemas, tables, views and functions; exact numbers, physical plans in Explain, and the warehouse’s queries in Server Activity.
Keyspaces, partition and clustering keys, indexes, types and functions; CQL with paging and consistency levels, grid edits saved as one batch, and query traces in Explain.
A region’s tables with their keys and indexes; PartiQL with items as columns and the capacity used, and grid edits saved as one transaction.
Each database’s keys by type, a key’s value as rows, the commands that make it again, replies as tables, and clients in Server Activity.
Indices, aliases and data streams with their fields; SQL or Kibana-style requests with hits as rows, the query DSL in Explain, and tasks in Server Activity.
Find databases in AWS, Google Cloud and Azure and add them with everything filled in; refresh a connection’s endpoint from the cloud.
YugabyteDB, TimescaleDB and AlloyDB Omni as PostgreSQL; TiDB and OceanBase as MySQL; FerretDB and Microsoft DocumentDB as MongoDB.
open existing or create missing file
HTTPS SQL API
Official driver; host and port or mongodb+srv:// connection strings, TLS and SSH tunnels.
DSN, driver, or connection string
configuration-file discovery
every database type
fills dialog from a PostgreSQL URL
Choose the context for your cloud data workflow.
Explore main and attached schemas.
Embedded analytics engine with Parquet and JSON readers built in.
A built-in DuckDB database that’s always there, for querying files and keeping results.
A database running in Docker on your Mac is matched by its port, with its state shown beside the connection.
One click from the connection’s menu; a refused connection offers to start its container and connects the moment the database is ready.
Works with each, and opens Docker Desktop for you when it isn’t running.
Tag each connection; Production switches on the Production guards and marks its consoles.
Hover a connection for its user, host, database, SSL mode or file path.
Connections, settings, saved queries, themes and keychain passwords come straight over.
PostgreSQL, MySQL and MariaDB on RDS and Aurora sign in with a fresh IAM token from your AWS profile.
Sign in with a profile of the Databricks CLI’s ~/.databrickscfg, host and token included.
Temporary credentials from your AWS CLI profile, SSO included, so there’s no database password to keep.
system keychain
default, environment, or chosen file
.my.cnf and applicable option groups
.mylogin.cnf
token/key passphrase can be stored securely
macOS GSSAPI ticket
includes a Test action
Use database trust or certificate authentication.
Configure secure connections using your database driver’s TLS options.
PostgreSQL, CockroachDB, Redshift, Materialize, CrateDB, RisingWave, ClickHouse, MySQL, MariaDB, StarRocks, Apache Doris, Oracle, SQL Server and MongoDB
agent can include 1Password
known_hosts
Select read-only session settings for supported database backends.
table-open and maximum-fetch limits
template-database option for PostgreSQL
principally PostgreSQL settings
Resolve jump-host passwords and key passphrases securely.
Identify jump-host and database connection status separately.
Asked on first connect and again if the server turns it down; keep it for the session only if you prefer.
PostgreSQL, MySQL and Oracle sessions carry the connection’s application name.
An UPDATE or DELETE without a WHERE asks first, and schema changes need the database name typed.
Run an INSERT, UPDATE, DELETE or MERGE in a rolled-back transaction and see exactly which rows it would affect.
On PostgreSQL and SQLite, Dry Run shows the rows before and after; Make Change then runs it for real.
On PostgreSQL, SQLite and SQL Server.
Nothing is committed until you say so; the toolbar shows when a transaction is open.
The database name is required on Production.
In Manual mode, submitted edits go into the console’s transaction behind a savepoint.
Every update must match exactly one row, or the whole batch is rolled back.
Kills, grants and user changes show their SQL first and are refused on read-only connections.
Shown as ******** in generated SQL, history, logs and error messages.
loads on expansion
Explore the objects provided by each database backend.
grouped under tables
Tick the ones to show, or hide one from its menu; a count like (2 of 5) by the connection opens the choice again. Oracle connections choose schemas.
Inspect database-specific object definitions.
can start without a connection
final result set displayed
dialect-aware splitting
Server-side cancellation through the database adapter.
Execution feedback and supported server notices in the Output tab.
Inspect identity/computed attributes and database comments.
Manage your sessions alongside database objects.
Drag a connection into place, or use Move Up / Move Down; the order is kept.
Open tables with dialect-appropriate SELECT SQL and configurable row limits.
Keep SQL Server temporary tables, SET settings, and USE context between runs.
Undo, redo, cut, copy, paste, and select all.
Find tables, views, columns, indexes, constraints, sequences, routines, triggers and types by name.
Find text in views, functions, procedures, triggers, packages, checks, column defaults and MongoDB validators.
Optionally case sensitive, in one schema, one database or every database on a connection.
Matches stream in with numbered, highlighted definition lines; double-click to jump straight there.
Open views, routines, triggers and Oracle packages as ready-to-run CREATE OR REPLACE statements.
⌘-click tables to compare two of them or diagram your selection.
One click on any SELECT.
Follow a foreign key to the row it refers to, or to the rows that refer to it.
Rows, run time, a live timer while a query runs, and totals of selected cells.
dialect-aware
automatic or explicit invocation
token validation before replacement
conservative scope-aware checks
Scope-aware name resolution across SQL statements.
aliases, CTEs, tables in statement
F2/context menu
Choose a folder; names, tags, descriptions, and connection context travel in comment headers. Individual SQL files make Git versioning straightforward: track changes, review diffs, and restore earlier versions of your queries.
Migrate your query library in either direction.
Bring your own query collection into the sidebar.
Pick up edits and Git pulls within seconds.
Match saved-query connections by name on another computer.
See new and modified queries alongside their branch.
Keep query work in version control without leaving the app.
Manage SQL and query metadata from the sidebar.
PostgreSQL function bodies, MySQL # comments, SQL Server [names], Oracle q-quoted strings and more.
Turn * into the real column names, qualified per table.
sel, ins, upd, cte, join and more, with Tab-through placeholders; add, edit and reset your own.
:name, ?, ${name} and $1 recognized automatically, asked once each, with last time’s values remembered.
Every statement you’ve run, grouped by day, with connection, rows, outcome and parameter values.
Open in a new console, insert at the cursor, run again or copy.
No auto-capitalization or autocorrect in any text box.
virtual scrolling
client-side sorting
case/whole-word/regex options
selection/editing follows layout
TSV/CSV options
floating or docked
Edit single-table query results containing the full primary key, or columns you choose with Identify Rows By. Neo4j nodes, Elasticsearch and OpenSearch documents, Redis values and InfluxDB fields are edited in their own terms. On Apache Doris, its UNIQUE KEY tables; on ClickHouse, saved as mutations it waits for.
Review highlighted pending changes before submitting the batch.
Work with full values in editable table results.
the full key, or columns chosen with Identify Rows By, must be in the results
Every update must match exactly one row; conflicting batches roll back together.
grid, copy/export, explorer
separate light/dark colors
Retain sorting, column widths, and scroll position on eligible read queries.
Reset the result area from its toolbar.
Fit headers and values automatically with practical limits for long text.
Save edits to the database identity associated with the result.
Respect the database’s identifier case rules.
Keep pending changes intact when using refresh and clear controls.
One tab per result set, each with its own edits, sort and chart; pin a tab to keep it across runs.
Tick values with their counts, match text, or Filter by This Value, instantly and without rerunning.
Count, sum, average, minimum and maximum in the status bar.
Syntax colors, Beautify and Compact, and editing.
On PostgreSQL and SQLite each column saves back to its own table.
New rows in green, deleted rows struck through, pasted rows added as needed.
Tab and ⇧Tab move to the next or previous editable cell.
database-specific SQL
Generate SQL appropriate for the connected database.
Configure database-specific keys and ON DELETE / ON UPDATE actions.
Database-specific index options
execute or open in console
Review the database-specific SQL before applying the change.
Copy rows and swap the rebuilt table in a transaction.
side-by-side differences
Recreate triggers transactionally with the updated table.
Plan related column and trigger changes in the table designer.
Keep explorer and SQL completion current.
Across connections, with a migration script in either direction and destructive steps safely commented out.
Rows matched by key, changed cells highlighted, differences exported as CSV or a sync script.
Keys, indexes, checks and rows, with types translated, even to and from MongoDB.
PostgreSQL, MySQL/MariaDB, SQLite, Oracle, SQL Server, Snowflake
cost/time-share bars and tooltips
PostgreSQL, CockroachDB, MySQL/MariaDB, SQL Server and Babelfish, Oracle, DuckDB, BigQuery, ClickHouse, StarRocks, CrateDB, Neo4j, InfluxDB, Elasticsearch, OpenSearch and MongoDB
Run analysis inside a transaction and roll it back afterwards.
Step through routines with the PostgreSQL debugger extension configured.
can create extension after server prerequisites
PostgreSQL, CockroachDB, Redshift, Materialize, CrateDB, RisingWave, ClickHouse, MySQL, MariaDB, StarRocks, Apache Doris, SQLite, DuckDB, Oracle, SQL Server, Snowflake and BigQuery (a dry run, with what a query would process).
Preview the effect of a change before making it.
Every session on the server, live: user, client, state, query and transaction age, waits and query text.
Blocked sessions and their blockers highlighted and drawn as a tree, with each session’s locks.
Every 2, 5 or 10 seconds, pausing while you work; filter, hide idle sessions or DataWorx’s own.
On PostgreSQL, MySQL, MariaDB, StarRocks, Apache Doris, SQL Server, Oracle, Snowflake and MongoDB; Cancel on Materialize, CrateDB, RisingWave and BigQuery’s running jobs. The statement is shown first.
Every user, login and role with attributes, memberships and privileges, held directly or through a role.
The complete privilege picture, including SQL Server DENYs.
See who can do what on any table, view, sequence, routine or schema, and grant or revoke right there.
New User, Role and Login, Edit, Add to Role and Drop, each with editable SQL before it runs.
Rows and table, index and TOAST/LOB sizes, index use with unused indexes flagged, and maintenance details.
Every table in a schema, largest first, with the total and each table’s share.
Without full privileges, every view shows what it can and explains what’s missing.
distinguished with dashed outlines
orthogonal relationship routing
separate Chart tab
columns, keys, indexes, and rows; types translated between databases; undone if it fails
UTF-8, UTF-16, Latin-1 fallback
Transactions handled by the connected database
typed cells, frozen header, filters, fitted widths
Format-appropriate clipboard output
add/duplicate/reorder/delete presets
PostgreSQL, MySQL, MariaDB, SQLite, SQL Server via SqlPackage, and Oracle via Data Pump.
up to 100,000 rows/run
references existing rows
Generate values for supported lists, ranges, and length rules.
up to 1,000 rows
Choose supported scopes in the database dump workflow.
Select supported output content options.
Custom, plain SQL, directory, and tar formats.
Use pg_restore/psql, mysql/mariadb, sqlite3, SqlPackage, or Oracle impdp as appropriate.
Choose the database command-line tools for your installation.
Manage long-running operations from a dedicated window.
Excel and Parquet keep their types.
Large files stream in; a 400,000-record file imports in about two seconds.
Realistic values chosen from each column’s name and type.
Columns such as ordered_at stored as text still get real dates and timestamps.
CSV, Parquet or JSON files, or whole folders read as one table.
Keep it as a typed table and join it with files and other results.
Rerun a pinned table’s query on its original connection.
Into the same or another database, created for you on PostgreSQL and MySQL, or another schema on Oracle.
Dump and restore SQL Server and Oracle with their own tools.
Dump and restore tools receive passwords securely.
dedicated choices, Gemini/Ollama use compatible APIs
one chat tab per enabled engine
up/down history
Connect through DataWorx --mcp using stdio.
hidden otherwise
per connection, max 5,000 rows and bounded response
internal chat and MCP
off by default, paths resolved inside allowed roots
Give assistant SQL work its own console context.
Inspect files and create folders within explicitly writable shared locations.
Keep file operations inside the roots you share.
Keep connection passwords out of assistant output.
Pick up shared connection changes without restarting the server.
The connection dialog shows the claude mcp add command to copy.
The DataWorx extension for Visual Studio Code runs on the same engine.
installed-font search
17 including Pastel Night
persisted locally
connection and data colors
in-window menus on other platforms
version, build, commit, email actions
JSON and environment-variable path overrides
Preview editor palettes and remove imported themes in Settings.
Toggle explorer, saved queries, AI, value viewer, and transposed results.
Open database documentation from the app.
View bundled component licenses and copyright details in About.
Save connections, preferences, query JSON, and themes using temporary-file replacement.
Keep the previous configuration version in a .bak file.
Preserve damaged configuration and restore an available backup.
Chosen separately for light and dark mode, with a live preview.
Choose the colors for strings, numbers, booleans and dates.
Window size and position, panel widths and layout restored exactly.
Every feature in one searchable page.
Every command with a shortcut, for your computer.
If a panel ever hits a problem it offers Try Again while the rest of the app keeps working.
A compiled desktop application with native platform integration.
Drag-to-Applications distribution.
Build tooling for Apple silicon and Intel.
Release tooling supports signing, hardened runtime, notarization, and stapling when configured.
Version, build, commit, and release-note stamping.
A test suite spanning application rendering, adapters, and workflow logic.
SQLite samples, container-based database setup, and mock AI/Snowflake services.
Native drivers for every database, tested against real servers.
License-generation tooling and searchable attribution inside the app.
Query MongoDB as you would in the MongoDB shell: find, aggregate, inserts, updates, indexes, show and use.
Collections, methods, fields sampled from your documents, and the $ operators that fit where you are.
Choosing an operator inserts its complete phrase; Tab moves from one placeholder to the next.
Every form of the operator or method you're writing, with the key you're on explained.
Fields follow the pipeline through $group, $project and $lookup; $lookup offers the other collection's fields.
Beautify long queries and see where a statement stops making sense.
Results open as MongoDB JSON; the table view shows sub-documents as JSON to edit in place.
Hover over an ObjectId in results, the JSON view or your query to see when it was created.
Edits keep each value's BSON type, are written by _id, and never overwrite someone else's change.
Browse each database's collections with the fields found in their documents.
a row per document, typed values, _id as the key, indexes kept
updateMany({}), deleteMany({}), drops and index changes ask first; read-only connections refuse every write, $out and $merge included.
Current operations in Server Activity with Kill, users and roles, and collection and index statistics.
Database-specific capabilities are named in the catalogue. This is the DataWorx product feature inventory; edition licensing is shown below.
REAL APP. REAL WORKFLOWS.
Captured from DataWorx running against sample PostgreSQL, DuckDB and MongoDB databases.
DEPTH YOU CAN PUT TO WORK
Robustness lives in the details: how changes are applied, how credentials are handled, and how your workspace is preserved.
Stage grid changes before saving. Each update must match exactly one row, and conflicting or failed batches roll back together. Primary-key matching and database-aware result identity keep updates grounded in the data you read.
The table designer previews database-specific CREATE and ALTER statements. SQLite rebuilds copy rows and swap tables transactionally, preserving triggers as part of the change. Explorer and completion metadata refresh afterwards.
System keychain storage, external password commands, pgpass, MySQL login paths, Snowflake tokens and keys, AWS IAM for Redshift, RDS and Aurora, your Google account or a service account key for BigQuery, Databricks CLI profiles, Oracle aliases, and SQL Server Kerberos support meet existing authentication workflows. SSH tunnels verify jump-host keys.
Enable MCP per connection, start with read-only access, and choose write permission explicitly. Results are bounded. Shared folders have separate permissions and symlink-aware root boundaries. Optional confirmation puts you in charge of assistant SQL execution.
Connections, settings, saved-query JSON, and imported themes use atomic saves. Previous versions are backed up automatically. Startup recovery preserves damaged files and restores available backups.
A fast native desktop shell and a dedicated database engine power the application. Engine tests, database integration scenarios, and mock service tests exercise its workflows. The screenshots on this page come from running the actual application against sample data.
BUILT TO STAND ALONGSIDE THE BIG NAMES
Database development, visualization, debugging, data generation, and AI in one $50 perpetual release license.
Table design, cross-connection structure comparison, ER diagrams, execution plans, and interactive PostgreSQL debugging sit alongside your SQL.
Transactional grid editing, charts, typed Excel exports, native dump/restore workflows, and realistic test-data generation belong in the same workspace.
Choose your provider or local Ollama model, give AI SQL/schema context, and share connections through MCP with per-connection permissions.
DataWorx brings advanced database workflows together for $50. Yes = available in the named product or edition. No = absent from that edition. ? = availability unverified. Comparable capabilities can differ in implementation.
| Workflow | DataWorx | DataGrip | DBeaver Community | DBeaver Enterprise |
|---|---|---|---|---|
| Dry run for changes, plus environment tagging | ✓ Dry run changes and tag environments | ? | ? | ? |
| SQL completion & formatting | ✓ Context-aware, dialect-aware | Yes | Yes | Yes |
| Environment tagging & production guards | ✓ Mark connections as production; block UPDATE/DELETE without a WHERE; require the database name before DDL; show a PRODUCTION tag on its consoles | ? | ? | ? |
| Database connections | Broad native SQL and NoSQL support + ODBC | Broad JDBC database support | Broad JDBC database support | Broad database support + ODBC |
| Built-in DuckDB engine | ✓ Query local CSV, Parquet, and JSON files directly; join PostgreSQL results with Snowflake results or local files | partial - no cross connection joins | partial - no cross connection joins | partial - no cross connection joins |
| MongoDB with native MQL | ✓ Shell-syntax consoles, pipeline-aware autocomplete, operator phrases and syntax help, JSON and table results | Yes | No | Yes |
| Native ODBC connections | ✓ DSN, driver or connection string; installed drivers found for you | No | No | Yes |
| Cross-connection SQL joins | ✓ Join PostgreSQL results, Snowflake results, and local files through DuckDB | ? | No | No |
| Editable data grids | ✓ Most database types; stage & submit | Yes | Yes | Yes |
| JSON/XML value viewing | ✓ Format, color, dock, or float | Yes | Yes | Yes |
| Native Excel (.xlsx) export | ✓ Typed cells, filters & frozen header | Yes | ? | Yes |
| Visual table design | ✓ Columns, keys, indexes & CHECKs | Yes | Yes | Yes |
| ER diagrams | ✓ Auto-layout, cross-schema links, SVG/PNG | Yes | Yes | Yes |
| Schema-aware test-data generator | ✓ Up to 100,000 rows per run | ? | No | Yes |
| Execution plans | ✓ Plan trees, cost highlights & Analyze | Yes | Yes | Yes |
| PostgreSQL routine debugging | ✓ Breakpoints, stepping, variables & stack | No | ? | Yes |
| Query-result charts | ✓ Bar, line, scatter & pie | Yes | No | Yes |
| CSV/TSV import & rich exports | ✓ Mapping, validation & typed output | Yes | Yes | Yes |
| AI-assisted database work | ✓ Multiple providers, SQL/schema tools | Yes | Yes | Yes |
| External AI access through MCP | ✓ Built-in stdio server; explicit opt-in | Yes | ? | ? |
| Native desktop implementation | ✓ Fast native Mac app, a 69 MB download | No | No | No |
| Free offering | ✓ Community edition | Non-commercial use | Community edition | Trial offerings |
| Paid release license | $50 USD · perpetual for the release | $109 USD / first year for individuals 2.18× DataWorx’s one-time price | Free Community product | $255 USD / year 5.1× DataWorx’s one-time price |
Additional checks: DBeaver cross-database joins; DataGrip runtime; DBeaver runtime. Native desktop implementation means a native compiled application rather than a Java runtime application. DuckDB JDBC connections are documented by both competitors; their bundled engine and combined file/result workflow remain unverified here. DBeaver cross-database joins are documented for Ultimate and Team Edition, outside the Enterprise column. An undocumented feature stays ? rather than being assumed absent.
Checked October 7, 2026 against official product documentation. The DataWorx column describes the product inventory. PostgreSQL debugging uses the server debugger extension; editable results use single-table data with a primary key or chosen identifying columns. The comparison uses DataGrip, DBeaver Community, and DBeaver Enterprise. No means the capability is absent from the named product or edition; ? means it is unverified. MCP availability is assessed for the current release, separately from documentation for upcoming releases.
VS. DATAGRIP
DataGrip offers a database IDE with JetBrains tooling. DataWorx brings completion, inspections, plans, diagrams, charts, and MCP together with PostgreSQL routine debugging and a $50 perpetual release license.
Official product overview ↗VS. DBEAVER
DataWorx brings query-result charts, native ODBC, and schema-aware test-data generation into its product toolkit. DBeaver Community does not include these workflows. Add PostgreSQL debugging and AI/MCP, and DataWorx offers substantial depth at $50.
Official edition comparison ↗VS. TABLEPLUS
TablePlus emphasizes native database editing and SQL workflows. DataWorx pairs a fast native desktop app with schema design, ER diagrams, plans, test-data generation, charts, and integrated AI/MCP.
Official product overview ↗VS. PGADMIN
pgAdmin focuses on PostgreSQL administration, including plans, ERD, and debugging. DataWorx combines PostgreSQL development workflows with MySQL, MariaDB, SQLite, SQL Server, Oracle, Snowflake, and ODBC in the same app.
Official feature overview ↗VS. NAVICAT PREMIUM
Navicat Premium covers multi-database development, modeling, AI, and data generation. DataWorx offers its own integrated combination of schema tools, diagrams, charts, test data, AI/MCP, and a $50 perpetual release license.
Official product overview ↗These comparisons describe product positioning and documented functionality. Explore the full catalogue above to see the depth of the DataWorx workflow.
SIMPLE BY DESIGN
Start with Community, or buy a perpetual license for the release. Your workspace, your choice.
FOR THE COMMUNITY
A welcoming way to get started with DataWorx and put your data to work.
FOR YOUR EVERYDAY WORK
A straightforward perpetual license for the DataWorx release you purchase.
More workbench. Less spending. DataWorx is $50 once. DataGrip’s individual commercial plan starts at $109/year — $59 more in the first year. DBeaver Enterprise is $255/year — $205 more per year than DataWorx’s one-time price.
Prices checked October 7, 2026, in USD before tax: DataGrip individual commercial pricing and DBeaver Enterprise annual pricing. DataGrip renewal discounts apply, and annual subscriptions include a perpetual fallback license. DataGrip also offers free non-commercial use; DBeaver Community is free.
Perpetual licensing applies to the purchased release. AI provider services are configured separately with your own account or local model.
A FEW GOOD QUESTIONS
You pay $50 once and can continue using the release you purchase. Future release offers and upgrade terms are announced separately.
PostgreSQL, MySQL, MariaDB, Microsoft SQL Server, Oracle, SQLite, DuckDB, Snowflake, MongoDB, and generic ODBC connections.
No. AI is available, but DataWorx will never force it on you. Use the SQL editor, data grids, schema tools, charts, and other database workflows independently. Enable an AI assistant only when you want one.
Yes. DataWorx supports OpenAI and compatible endpoints, Anthropic, Gemini, and Ollama. You configure your own provider credentials or local model.
Yes. Store saved queries as SQL files, organize them with tags and descriptions, and use a Git-backed query folder to keep your work together.
DataWorx includes macOS native menus, standard shortcuts, system keychain integration, and light, dark, or system appearance.
MAKE ROOM FOR YOUR NEXT IDEA