ALog is an open-source, AI-powered audio diary app for iOS, hosted on GitHub with about 460 stars under the GPL-2.0 license and updates in 2026. Instead of typing journal entries, you speak them, and AI turns your voice notes into an organized diary.
The app is distributed on the App Store, and its full source code is published so anyone can build and run their own version. Spoken entries are sent to OpenAI models through a small backend that the project deploys as a Cloudflare Worker, which keeps API keys off the device and lets self-hosters control the AI model used. The iOS project is generated with XcodeGen and uses Arkana for secret management, and the repository documents the complete setup, from cloning and installing Ruby gems to configuring the worker’s environment variables such as the OpenAI key and optional HMAC validation.
The design reflects a simple observation: people will keep a diary they can talk to far more reliably than one they must type into. Entries start as quick voice recordings, and the AI layer handles the tedious part, transcribing the speech so the journal remains readable and searchable later. Because the backend is a self-deployable Cloudflare Worker, technically inclined users can run the whole system under their own accounts.
ALog is free and open source under the GPL-2.0 license, and the App Store version is free to download. If you self-host the backend, you only pay for your own OpenAI API usage.
It is aimed at journalers who think out loud and want a diary that keeps up with them, and at iOS developers interested in a complete, open example of a voice-driven AI app with a serverless backend.