LEETLABS
§ 00 / FIELD NOTE
BY Said Betmurzaev · Founder, Leetlabs

Retool alternatives: when to buy, and when to build

Most people searching for a Retool alternative aren’t unhappy with Retool. They’ve hit one of three walls: the per-seat bill climbs every time someone new needs access, a workflow outgrew what a drag-and-drop canvas can express, or a security review flagged that the app logic now lives inside a vendor they don’t control. Which wall you hit decides whether the answer is another tool or your own code.

Retool sits at the center of this market for a reason. It raised a $45 million round at a $3.2 billion valuation in July 2022, and the product genuinely earns its place for admin panels and CRUD dashboards over a database. The question isn’t whether it’s good. It’s whether an internal-tool builder is the right shape for the thing you’re actually building.

The open-source alternatives, and the license fine print

If the wall you hit is cost or control, three source-available builders cover most of what Retool does, and you can self-host all of them:

  • Appsmith is licensed under Apache-2.0, the most permissive of the three. You can modify it, run it internally, and embed it in a proprietary product without publishing your changes.
  • ToolJet ships under AGPL-3.0. That’s a strong copyleft. If you expose a modified ToolJet to users over a network, the AGPL’s network clause obliges you to offer them the modified source.
  • Budibase uses GPL-3.0, copyleft without the network clause.

The license is the part teams skip and later regret. Apache-2.0 clears legal review quickly and is safe to embed. AGPL and GPL are fine for internal use, where you aren’t distributing anything, but they change the calculus the moment you want to bundle the tool into something you sell. Read the license before you standardize on a platform, not after.

Where every internal-tool builder stops being the answer

Here’s the pattern we see when we’re brought in to replace one of these tools. It’s rarely a cost decision. It’s that the tool did its job well enough that the business started asking it to do a job it was never built for.

The tell is custom logic. Internal-tool builders are excellent at reading and writing rows. They get awkward fast around anything stateful or bespoke: multi-step approval flows with branching, real permission models that go past table-level access, background jobs, webhooks that have to be idempotent, or a data model that doesn’t map cleanly to tables and forms. You end up writing more and more JavaScript inside the vendor’s editor, in a sandbox you can’t test properly, version-control cleanly, or run locally. At that point you’re already building software. You’re just doing it inside someone else’s constraints, with worse tooling than a normal codebase gives you.

The other tell is ownership. A tool your operations run on that lives entirely in a hosted low-code platform is a dependency you can’t audit line by line and can’t take with you. For a throwaway internal panel that’s a fine trade. For a workflow that becomes load-bearing, it’s a slow-building liability.

A rule of thumb that has held up

Buy when the tool is a means to an end and stays that way: a dashboard, an admin CRUD panel, an occasional data-entry screen, something a handful of internal people touch. The value is in having it today, not in owning it.

Build when the tool becomes the product, or a real part of it, when it needs a permission model you’d have to explain to an auditor, or when the roadmap keeps adding branching logic that fights the canvas. The signal is simple: if you’re writing significant code inside the low-code editor to make it behave, you’d be better off writing that code in a real codebase you own.

That “build” path is what our MVP development work exists for. We ship a first version that’s deliberately small but real underneath, deployed and running on code you keep, rather than a prototype trapped in a platform you’ll have to escape later. The goal is that the thing proving your idea is also the thing you keep building on.

If you’re on Retool today and it’s working, stay. The moment to move is when you notice you’re fighting the tool more than the problem. That’s not a signal to shop for a different builder. It’s usually a signal you’ve outgrown the category.