dbForge Studio vs SQLly: the Windows SQL Server suite and a cross-platform workbench
An honest look at dbForge Studio for SQL Server next to SQLly: editions and prices, Windows-only vs native on three OSes, where dbForge's depth wins, and where SQLly differs.
dbForge Studio for SQL Server is the tool a lot of SQL Server teams buy when SSMS stops being enough. Devart has been building it for many years, and it shows: schema and data compare, source control, a T-SQL debugger, tSQLt unit tests, a profiler, a monitor, a test data generator, a documenter and CI plugins, all in one Windows app. The current release is 2026.2.82, from 16 October 2026.
We build SQLly, so take the bias into account. For this post we read Devart’s product, editions, ordering and download pages and its installation docs. We did not install dbForge Studio or benchmark it, and we didn’t try it under CrossOver. On SQL Server depth it is ahead of SQLly in several places, and we say which ones.
The family
dbForge is a line of products, not one app. There is a Studio each for SQL Server, MySQL (which also covers MariaDB), PostgreSQL and Oracle, and dbForge Edge bundles all four. This post is mostly about the SQL Server Studio, since SQL Server is where both tools put the most work.
Platforms
Devart describes dbForge Studio as “a Windows-native application”. It is a .NET app that needs .NET Framework 4.7.2 or later. On macOS and Linux, Devart’s docs walk you through CrossOver, the CodeWeavers compatibility layer built on Wine: create a bottle, install .NET Framework 2.0, 2.0 SP2, 4.0 and 4.7.2 into it with the CrossOver installer, then install the Studio. The docs list CrossOver 21 as the prerequisite and mention Parallels Desktop as the alternative. The Windows installer is 140.09 MB.
SQLly is one Rust codebase drawn on the GPU through Metal, Direct3D or Vulkan, and it runs natively on all three. The macOS build is signed and notarized and needs Apple Silicon; Windows and Linux are a supported beta on x64 and arm64. The Windows zip is 60.9 MB and the Mac dmg 65.6 MB.
If everyone on your team is on Windows, this section matters little. If half of them are on Macs, it matters a lot.
What it costs
dbForge Studio for SQL Server comes in four editions:
| Edition | Subscription | Perpetual | Adds |
|---|---|---|---|
| Express | Free | Limited completion, formatter, designers, diagrams, data editor, query builder, backup/restore, Monitor | |
| Standard | $249.95/yr | $499.95 | Debugger, Query Profiler, snippets, refactoring, pivot and reports |
| Professional | $399.95/yr | $799.95 | Schema and data compare, test data, code analysis, Events Profiler, command line |
| Enterprise | $549.95/yr | $1,099.95 | Source control, unit tests, documentation, script-folder compare, DevOps plugins |
The subscription prices are on Devart’s editions page; the perpetual ones come from its ordering page, where a perpetual license includes an initial support and upgrade period you can extend. Two- and three-year subscriptions are discounted, priority support adds 20%, and there are volume discounts from five seats. A new install runs as Enterprise for a 30-day trial and then drops to Express. dbForge Edge, the four-Studio bundle, is $749.95 a year or $1,499.95 perpetual.
SQLly has a free tier for hobbyists (two saved servers, one cloud account, a reminder to pay about every five hours, not for organizations over 100 employees), a $100 one-time SQLly 1.x license, and a $200 Lifetime license. See pricing. A free Express edition at any company size is something dbForge offers and SQLly doesn’t; we couldn’t find Express’s exact license terms on Devart’s site.
Where dbForge is ahead
This is the honest list, and it’s not short.
- T-SQL debugging. dbForge has a debugger from Standard up. SQLly has none.
- Source control for the database itself. Enterprise links objects to Git, TFVC, SVN, Perforce, Vault, Mercurial and Azure DevOps, shows object history, and compares against script folders. SQLly has a git panel for script folders and can overlay local DDL files on a live schema, but it does not version database objects.
- Unit tests. Enterprise runs tSQLt tests. SQLly doesn’t do unit tests.
- CI. Command-line compare, data generation, import/export and documentation from Professional, plus PowerShell cmdlets and Jenkins, TeamCity, Azure DevOps and Bamboo plugins in Enterprise. SQLly’s dialogs can copy a
sqlly-clicommand for exports, backups and dumps, which is far less. - Plans and profiling. The Query Profiler shows plan diagrams and wait stats and compares runs side by side. SQLly’s plan view is still marked in progress.
- Documentation. HTML, PDF or Markdown documentation in Enterprise. SQLly writes a Markdown data dictionary and nothing fancier.
- Microsoft platform coverage. Devart lists SQL Server back to 2000 (with limits), Synapse pools, Fabric Data Warehouse and OneLake. SQLly doesn’t document any of those.
Where the two overlap
A good part of the daily work is covered by both. Both have code completion, a formatter, visual table designers, ER diagrams, security editors for logins, users and roles, a visual query builder, backup and restore, a test data generator, schema and data compare, and a server monitor. dbForge’s Monitor shows CPU, waits, top queries, sessions and data I/O; SQLly’s Activity Monitor shows sessions, blocking chains, expensive queries and waits, next to a Server Dashboard, Query Store insights and Extended Events. The details differ, and dbForge puts compare and data generation behind Professional.
Where SQLly is different
- One app for 30 engines. SQL Server and Azure SQL sit next to PostgreSQL, MySQL, MariaDB, Oracle, SQLite, DuckDB, ClickHouse, libSQL and Redis in the same window, under one license. The engine list says which tools work where.
- IntelliSense that reads your data. SQLly pulls a column’s real values into a
WHEREclause, writesJOIN ONfrom actual foreign keys and flags aGROUP BYmistake while you type. IntelliSense - Guardrails. SELECT-only connections that block file-reading and shell-escape tricks, a hard stop on
UPDATE/DELETEwithoutWHERE, auto-rollback, tenant checks and production painted red, inherited from client to project to environment. dbForge has connection coloring and, in Enterprise, execution warnings. Safety - Big results. SQLly spills to disk past 50,000 rows or 64 MiB, so 50 million rows use about as much memory as a small result. We couldn’t find how dbForge handles very large results.
- AI on your terms. dbForge’s AI Assistant uses OpenAI and Anthropic models through Devart, sends schema metadata rather than data, and is a separate license: 200,000 tokens a month free, or 20 million for $9.95 a month or $99.95 a year. SQLly has no default provider; you choose a local model (Ollama, MLX), your own endpoint, or an opt-in cloud provider.
- Agents that only read.
sqlly-cli mcplets an agent read schemas and run oneSELECTin a read-only transaction that is always rolled back; writes come back to you as proposals. Devart sells separate MCP Servers, but we found none built into dbForge Studio.
Which one to pick
Pick dbForge Studio if:
- Your team is on Windows and works almost only in SQL Server.
- You debug T-SQL, run tSQLt tests, or keep database objects in source control.
- You deploy schema changes from Jenkins, TeamCity, Azure DevOps or Bamboo.
- You read execution plans every day and want a finished profiler.
- You want a free tool and Express’s limited feature set covers what you do.
Pick SQLly if:
- You or your teammates work on macOS or Linux and want a native app, not a compatibility layer.
- You work across SQL Server and other engines - PostgreSQL, MySQL, Oracle, DuckDB, ClickHouse, Redis.
- You want guardrails on your own queries and agents that can read but never write.
- You want AI that can run on your own machine.
- You’d rather pay $100 or $200 once than a yearly fee.
The fair test is your own database. The side-by-side table is at SQLly vs dbForge Studio, and SQLly’s download is free with no account.