Frequently Asked Questions

ANDROID-ARCHITECT is an AI-powered platform for planning Android apps. Describe an app idea in plain language and Google Gemini writes an architecture blueprint — feature list, tech stack, a Mermaid architecture diagram and a Play Store listing draft. From that blueprint it generates, on demand, Room and Retrofit data layers, Kotlin feature code, CI/CD configurations, a design system and AI code reviews.

Create a free account — no credit card is required. Open Architect, describe your app under App Vision, choose a Technology Core and press Architect. A compliance check runs first, then the blueprint is generated. From its tabs you can generate the data layer, CI/CD configuration, feature code and design system. Test Drive loads a sample blueprint instantly if you want to look around first.

No. Every account runs on included AI credits, so you can describe an app idea and get an architecture blueprint without supplying a key. Bring-your-own-key — running generations on your own Google Gemini quota instead of ours — is in development and is not available yet.

Some Android knowledge is helpful but not required. The platform is designed to accelerate developers at any level. Describe your app idea in plain language and the AI generates an architecture blueprint and first-draft code you can build from. Experienced developers use it to compare architecture options in the Architecture Advisor and to draft data layers, pipelines and feature code quickly.

The Blueprint Builder (Architect in the sidebar) is the workspace where a blueprint is generated and worked on. You describe the app and pick a technology core; the AI writes the features, tech stack, a Mermaid architecture diagram and a strategy document. Its tabs then generate the data layer, API client, CI/CD configuration, feature code, design system and store listing from that blueprint. Features cannot be edited by hand yet, and there is no architecture-pattern selector — use the Architecture Advisor to compare patterns.

The AI Chat is a streaming conversation with Google Gemini. Ask about Android architecture, Kotlin patterns, Jetpack Compose, dependency injection, testing, or any Android development topic. It keeps the conversation history for follow-up questions; it does not read your blueprint, so paste in the details you want it to consider.

Kotlin: Room entities and DAOs with an entity-relationship diagram, Retrofit API clients, per-feature implementation code, and — from the Dev Tools tab — Compose screens, ViewModels, repositories, navigation and Hilt modules. It also writes CI/CD configurations, and the Test Generator tool on Pro writes unit and UI tests. Each piece is generated separately from your blueprint; the platform does not assemble a complete, buildable project.

Yes. The Template Library provides pre-built app architectures for common use cases including e-commerce, social media, fitness tracking, and productivity apps. Each template includes a full feature set, tech stack configuration, and architecture blueprint that you can customize in the Blueprint Builder before generating code.

The Blueprint Builder's Deploy tab generates GitHub Actions or Bitrise configurations. The CI/CD Generator tool on the Pro plan covers six providers: GitHub Actions, GitLab CI, Bitrise, Jenkins, CircleCI and Azure DevOps, with options such as lint, test, signing and Detekt or Ktlint steps. The configurations are AI-generated, so review them before you commit them.

ANDROID-ARCHITECT uses Google Gemini for its AI features, including blueprint generation, code generation, compliance analysis, and the chat consultant. Most generators request structured JSON output against a schema, so their results arrive in a consistent shape; chat and feature code are free-form.

Treat it as a first draft. The code is written by an AI model and the platform does not compile, run or test it, so review, test and adapt it before you ship anything. The QA & Test tab runs an AI code review of generated feature code and suggests fixes, which helps you find problems faster but is not a substitute for building and testing.

Before generating, you control the description and the technology core (Kotlin with Compose, Kotlin with XML views, React Native or Flutter). After generating, everything is yours: copy or export it and change it freely. The output is standard code with no proprietary runtime or lock-in. Editing a blueprint's features in place is not supported yet — change the description and generate again.

Blueprints target Android. The Multi-Platform tool on the Pro plan can also generate Kotlin Multiplatform, iOS (Swift and SwiftUI), Flutter and React Native code. There is no desktop target.

Yes. The Starter plan is free and needs no credit card. It includes 50 AI credits a month, the Blueprint Builder, AI Chat, the design system generator and the template browser. The Pro tool panels need the Pro plan. See the Pricing page for the full comparison.

Pro ($29 a month, or $24 a month billed annually) adds 500 AI credits a month, a larger project allowance, unlimited exports, and the Pro tool panels — CI/CD, security, testing, performance and release tools, plus the Architecture Advisor and voice commands. Team plans with collaboration have not launched yet. The Pricing page is the authoritative list.

Not yet — it is in development. Until it ships, every generation runs on our Gemini key and is charged against your credit allocation. A key saved under Settings is stored encrypted in our database and is not currently used for generation, so adding one today changes nothing about how your account works.

Yes. Starter includes 50 AI credits per calendar month, reset on the 1st. Each AI action costs a fixed number of credits: a chat message 1 credit, a design system 3 credits, a data layer 5 credits, feature code 10 credits (5 for the code plus 5 for the AI review that runs on it automatically), and a new blueprint 13 credits — 10 for the blueprint plus 3 for the compliance check that runs first. Starter also has a project allowance, 3 exports a month, and no access to the Pro tool panels.

Your account, subscription and usage records are stored in Supabase behind row-level security, and payments are processed by Stripe. The prompts you send are processed by Google Gemini to produce each result. We do not sell your data. Blueprints you create while signed in are stored in your account as projects, and you can delete them from the Dashboard. See the Privacy Policy for processors, retention and your rights.

Generated code is returned to your browser and shown in the interface. Your blueprint is saved — in your browser, and to your account as a project when you are signed in — so a reload does not lose it. Generated code files are not stored on our servers unless you push them to GitHub. Export a blueprint as PDF, JSON or ZIP to take it elsewhere, and manage code in your own version control.

It is built with the GDPR controls a user needs: cookie consent with separate categories, respect for Global Privacy Control (GPC) signals, and account deletion you can do yourself from your profile. Change your consent at any time from the Cookies link in the footer. The Privacy Policy sets out the legal basis, the processors and your rights.

Email and password, Google, or GitHub — all handled by Supabase Auth. AI features require you to be signed in.

Use a current version of Chrome, Edge, Firefox or Safari with JavaScript enabled. The workspace is designed for a desktop screen; the architecture diagrams and code views need the room.

The interface is responsive and accessible on tablets and mobile devices. However, the Blueprint Builder and code generation features are designed for desktop use where you benefit from larger screen real estate for architecture diagrams and code review. Mobile works well for the AI Chat consultant and documentation.

Partly. A service worker caches the app, so previously loaded pages open without a connection. Every AI feature — chat, blueprints, code generation — needs a connection, because it runs on Google Gemini.

Not yet. Starter and Pro are single-user plans. Sharing and collaboration are planned for a Team plan that has not launched — use Notify Me on the Pricing page to hear when it does. Until then, share a blueprint by exporting it as PDF or JSON.

Not yet. The plans on sale are Starter and Pro, and there is no single sign-on. If you need more than Pro provides, email [email protected].

Not yet. Real-time collaboration is planned for the Team plan, which has not launched. Today each account works on its own blueprints.

Generation time depends on the size of the request; a blueprint is the largest. If a generation fails, first check that you are signed in and have credits left — the balance is shown in the sidebar. When credits run out, AI features stop until the 1st of the next month or until you upgrade. Otherwise check your connection and retry, or split a large request into smaller ones.

Email [email protected]. Include what you were doing, what you expected to happen, and any error message you saw. Feature requests are welcome.

Because bring-your-own-key is not available yet. A key saved in Settings is stored encrypted, but nothing uses it: every generation runs on the platform's Gemini access and is charged in credits.