A Bitrig alternative with no daily cap

It is worth stating how much these tools have in common, because the marketing in this category tends to obscure it.

Both are macOS apps. Both drive real Xcode and produce real Xcode projects. Both moved to bring-your-own-agent, so you authenticate with a Claude or ChatGPT subscription you already pay for rather than buying inference through a vendor — Bitrig through the Agent Client Protocol with Claude Code, Codex, Cursor, Grok Build and OpenCode; Modaal with those and several more. Both let you export. Both now have a free tier, and both keep App Store distribution for paying customers.

Both are also right about the same underlying thing: a native app should be a native project you own, not a build artifact you rent. If you are choosing between these two, you have already made the important decision.

This is the difference that shows up first, so it is worth being concrete.

Bitrig Free gives you "10 daily requests using your Claude or ChatGPT subscription" and "15 daily credits (up to 60 per month)". Requests run against the AI subscription you are already paying for. Credits are Bitrig's own, consumed "based on the complexity of the task."

Ten requests is somewhere around an hour of real work. Not an hour of demo — an hour of the actual loop, where you ask for a screen, look at it, ask for a change, hit a build error, ask about the build error, and go round again. Debugging is where prompts disappear fastest, and debugging is also where a cap hurts most: you start rationing attempts at exactly the moment you need to try things freely.

Modaal Free has no prompt cap. Not a high one — none. You can spend a Saturday on it and never see a limit.

We can do that because of how the product is built rather than out of generosity. Modaal does not buy inference. You bring your own agent and authenticate with your own subscription, and Modaal orchestrates it locally against Xcode on your own machine. A free user costs us very little, so there is no reason to meter them. Bitrig bundles its own credits alongside your subscription, which is a legitimate design — it just means someone has to pay for those credits, and on the free tier that is done by capping you.

Modaal Free is one active project. If you want to keep three apps going at once without paying, Free will not do it and Pro is €9 a month billed annually.

Bitrig's pricing page does not state a project limit on its free tier. We are not going to guess at what that means — if running several projects for free matters to you, go and check theirs. It is a fair question and we would rather you got an accurate answer than a flattering one.

What we would say is that one active project is the shape most people actually need while evaluating a tool. You are trying to find out whether this thing can build your app. One project, unlimited prompts, real native output and code you can take away is a genuine answer to that question.

Bitrig builds for Apple platforms and only Apple platforms. That is not an oversight — it is a coherent product decision, and it is part of why their Apple coverage is as deep as it is. Focus buys you something real.

But it means that if Android arrives on your roadmap, Bitrig cannot follow. You start a second codebase, in a second tool, with a second workflow, and then spend your time reconciling two products that were never designed together.

Modaal Free lets you pick either platform. Pro builds both together, from a shared core, with native Kotlin and Compose on Android and Swift and SwiftUI on iOS — so the business logic is the same on both sides while each app follows its own platform's conventions. If you are honestly unsure whether Android will ever matter, that uncertainty is itself the argument.

Both tools produce good code early. The question that decides whether a project survives is what the codebase looks like once an agent that does not remember last week has added fifty features to it.

Modaal answers this explicitly. The production template scaffolds a RIBs module tree — Uber's architecture pattern — with an explicit module hierarchy, protocol boundaries between every module, and a navigation tree that is explicit in the router hierarchy. Because the structure is on disk rather than in a conversation, it survives context clears, agent switches and a three-week gap. The quick-prototype template uses MVVM instead, deliberately smaller, keeping view models unit-testable.

Worth knowing before you start: migrating from MVVM to RIBs is a full rewrite, not a refactor. Choose the production template if the app might become real.

Bitrig's documentation does not describe an equivalent architectural commitment. That does not mean the code it writes is bad — given who built it, it very likely is not. It means the guarantee is different in kind: theirs rests on the quality of the generation, ours on a structure the agent has to write into.

We would rather you got this from us than found it out afterwards.

Platform depth. Widgets, Live Activities, Controls, Shortcuts, Siri, Spotlight, CloudKit, tvOS, visionOS, macOS. If your app lives in those corners, Bitrig has more of the surface covered.

Managed submission. Archives, code signing, upload, listing management and screenshot capture, handled for you.

Agentic UI testing. Bitrig tests the apps it builds by tapping, scrolling and typing like a user. Modaal runs a build-and-fix loop through Xcode and scaffolds a passing test suite, which is a narrower thing.

Bitrig Remote. Driving a build from your phone is genuinely novel and we have nothing like it.

The founders. Kyle Macomber and Jacob Xiao co-created SwiftUI. Tim Donnelly has an Apple Design Award. That is not a marketing line and it is reasonable to weigh it.

If those matter more to you than an uncapped free tier, Android, and a €9 Pro plan, Bitrig is the better choice and you should take it.