Aila (AISuperDomain): a .NET MAUI desktop shell that puts 20+ chat AIs in one window

What Aila solves, and who it is actually for

Asking the same question of five chat services means five tabs, five logins and five different scroll positions. Aila collapses that into a single desktop window: you type once, and the app displays answers from several models at the same time. The README frames this as getting "a broad spectrum of insights" from each model's distinct perspective. The practical audience is narrow and specific. It is for a person who already has accounts on ChatGPT, Gemini, Claude, Copilot, Poe, Perplexity, Mistral, Grok, Kimi Chat, Tongyi Qianwen, DouBao and the rest of the listed services, and who wants to compare their answers interactively. It is not for a backend service, a CI job, or anyone who needs a stable machine-readable response. Aila has no API layer in the repository listing; the top-level files are XAML pages, C# configuration classes and a config.json. The project is written in C# on .NET MAUI and ships as a desktop application for Windows and macOS, with the repository description also mentioning Android.

How the WebView layout and config.json fit together

The architecture visible in the repository is a MAUI shell around embedded browser views. WebViewManager.cs and CurrentAi.cs handle the browser surfaces, AppShell.xaml and MainPage.xaml define the layout, and ConfigurationManager.cs reads settings. The display count is a real constraint, not a cosmetic one: the README offers 1, 2, 3, 4 or 6 simultaneous panels, so six is the ceiling before you start switching. Interaction is keyboard-driven. Typing "/" opens a list of prompt suggestions, and a leading "#" marks a prompt as persistent so it stays available across submissions. Tab+Enter submits, and a Switch button cycles through the displayed models. Clicking an AI's name in the top-right corner opens that answer full screen, and a Home button returns to the grid. Everything about which models appear, and what prompts they receive, lives in Aila's JSON configuration file, reachable from the fly-out menu behind the top-left icon. That file is the extension point: the README says users can add new AI models there, modify prompts, and personalise the experience. The README also describes a script-based path for adding an AI that is not supported: inspect the input box element with the browser devtools, hand the HTML to ChatGPT, and ask it to write a generic script. That is the honest shape of the project. It is a configurable wrapper, and the wrapper's correctness depends on the pages it wraps.

Installing Aila on Windows and macOS

There is no package manager step and no build-from-source instruction in the README. The documented path is to download the latest version from the GitHub releases page. On macOS, the README says to double-click the downloaded file and follow the on-screen instructions. On Windows, two prerequisites come first: .NET 8 must already be installed, and you must install the certificate the project provides before running the Aila installer. Skipping the certificate is the most likely first-run failure. After installation, the README points you at the settings to customise your AI experience.

The releases listed in the repository are v1.1 from 2024-03-24, v1.0 from 2024-03-10 and v0.1 from 2024-03-08, so the published installers are older than the code on the main branch. If you want the current state of the source, you need the .NET MAUI toolchain rather than the release download. Once Aila is running, the first real action is to open the configuration file from the fly-out menu and confirm which models are enabled. The README does not document a schema for that file, so treat the shipped config.json as the reference and edit conservatively.

Where Aila breaks, and when it is the wrong tool

The failure mode is structural. Aila drives AI websites through embedded browser views, so any change to a site's markup, login flow or bot detection can break a panel without warning. The README's own troubleshooting advice for an unsupported AI is to inspect the HTML and have ChatGPT write a replacement script, which tells you the maintenance burden sits with the user. There is no documented health check, no fallback when a WebView fails to load, and no statement about what happens to a panel when a session expires. The README is also silent on data handling: it does not say whether prompts or responses are stored, logged or transmitted anywhere beyond the AI sites themselves. If your input is confidential, that silence matters. Aila is the wrong tool for anything programmatic. There is no documented command-line interface, no headless mode, and no stable response format, so you cannot diff outputs between models across runs or wire the result into a pipeline. And the README documents Windows and macOS installation only; the repository description mentions Android, but no Android install steps appear in the text.

Aila versus calling the model APIs directly

The obvious alternative is to skip the wrapper and call each provider's API from your own script. The difference is not convenience, it is what you get back. An API call returns a structured response you can store, compare and re-run, and it does not depend on a web page continuing to look the way it looked yesterday. Aila gives you the vendor's own chat interface, including whatever features that interface has and the API does not, at the cost of that interface being a moving target you do not control. A second alternative is a browser with several pinned tabs, which costs nothing and breaks no more often than Aila does, but gives you no shared prompt box, no six-panel grid, and no persistent prompt mechanism. Aila's value is concentrated in the prompt-once-see-many workflow and the JSON config that lets you define which models participate. If you do not need that specific interaction, the wrapper adds a dependency without adding capability.

Maintenance, licensing and what the repository tells you

The repository is not archived and the last push was on 2026-08-09, so the code has moved since the v1.1 release in March 2024. The files v2.0.txt through v2.3.txt sit at the top level alongside update.txt, which suggests version notes are kept in the repository rather than only in release entries. That is a signal about where to look for changes, not a promise of cadence. The licence is Apache-2.0, which permits commercial use and modification and requires that you keep the licence and notice files and state significant changes; it also includes an explicit patent grant. That is a summary of the licence text, not legal advice, and if you plan to redistribute a modified Aila you should read LICENSE yourself. The upgrade cost is uneven. Installing a new release is a download and a double-click. Keeping a custom AI script working is ongoing work, because it is tied to another company's page structure. Budget for the second cost, not the first.

Editorial conclusion

Aila fits people who already hold accounts on several chat services and want to compare answers without juggling browser tabs, and who accept that the app drives those sites through embedded WebViews rather than official APIs. It is the wrong tool if you need reproducible API output, unattended batch runs, or a Linux build, since the README lists Windows and macOS only and the last push to the repository was on 2026-08-09. Verify two things before relying on it: that the AI you care about still loads correctly inside the WebView, and that the JSON entry for that AI still matches the current page markup, because the README's own customisation path assumes you will fix the script yourself when a site changes.

Frequently asked questions

Official sources

1. License: Apache-2.0
2. Project website
3. README
4. Releases
5. win4r/AISuperDomain on GitHub