Flockbay for studios, with the caveats up front

No team plan. No per-seat licensing. No admin console, no role permissions, no single sign-on. Flockbay is a desktop application that one person installs on their own Windows machine and signs into. Two people can share a game by invite; that is not a rented workspace and not a seat.

If your procurement process needs a vendor with all of that, we are not that vendor today and pretending otherwise would waste a meeting. Flockbay is free on your own AI: your Claude or ChatGPT plan, your own agent, or an API key.

Everything below assumes you are fine with people bringing their own AI rather than buying a floor licence.

More on this: Game maker for schools: the deployment questions

The expensive thing in a studio is not building the prototype, it is deciding whether the idea is worth building. That decision usually waits on an engineer being free, and by the time one is, the question has gone cold.

Flockbay collapses that. A designer builds the world by hand and describes the behaviour; Flockbay writes the GDScript; the thing runs locally on the bundled Godot and someone plays it that afternoon. The output is a real Godot project, not a mockup, so what you are judging is a game rather than a slide.

That is a real reduction in the cost of finding out an idea is bad, which is where most of the value in prototyping lives.

Writers, artists, audio people and producers usually have ideas they can only express by asking someone else to build them. That request has a queue and the queue is long enough that most of the ideas die in it.

Given this tool, those people can build a playable version of what they mean. Not shippable, and not intended to be. Playable, which is a completely different quality of argument in a review meeting.

They also judge the result themselves, which is worth saying plainly: not writing the code has never stopped anyone knowing whether a game feels right.

Production on a large existing codebase. If you have a mature Godot or Unity project with an established pipeline, this is not a tool that plugs into it. It is app-shaped, not plugin-shaped.

Two people can share a game by invite. One holds it at a time; the other plays the latest version locally and can take over when it is free. There is no simultaneous editing, no version-control workflow, and no live shared world. What they make stays an ordinary Godot project on each machine.

And building requires an internet connection, because the chat, asset generation and world data come over the network. If your studio machines are on an isolated network, that is a hard blocker and there is no offline mode.

Desktop app for Windows. Bundles Godot. Games run locally at the machine's framerate — no streaming, no cloud build, no browser preview. Projects are ordinary Godot projects, which means nothing built here is stranded if you move the work into your own pipeline afterwards.

An external agent can drive the app over MCP if your team already has its own AI tooling and would rather use that than Flockbay.

More on this: Ziva alternatives, and why Flockbay is not one

Put it on one machine belonging to one person who has an idea they cannot currently build. Give them a fortnight and one question to answer: is this idea worth an engineer.

Do not roll it out, do not write a policy, and do not have a meeting about tooling. The whole point is that the cost of finding out is someone's spare afternoons and the AI they already have, and any process wrapped around that immediately costs more than the thing it governs.

If the answer comes back playable and interesting, you have a prototype and a decision. If it comes back dull, you saved a quarter.