Tinariox — a Local AI Companion for Android: Beta with Voice and Memory
Published: 2026-10-11 · Author: AI Release · @ai_release1
⚡ The gist in 5 seconds - A beta version of Tinariox has been released — a local AI companion for Android. - It connects to Ollama, LM Studio, or any OpenAI-compatible cloud endpoint. - It works with voice input, memory, and tasks right on the device. ### 🔍 What was found A beta APK has been published in the SevernyiVeter/tinariox-app repository. It's a free local AI companion for Android: it lets you chat with an LLM using either local servers (LM Studio, Ollama) or OpenAI-compatible cloud APIs. Built-in voice input, memory, and task features are advertised. How it works: the app follows a local-first paradigm, meaning data and processing stay on the device, while the model is connected via a custom endpoint. For local use, you can run Ollama or LM Studio on a PC and enter its address in the settings; for the cloud, you can specify any OpenAI-compatible URL. Voice, memory, and tasks are implemented on the app side, without mandatory calls to third-party services. ### 💡 Why it matters Tinariox is another step toward private AI assistants that don't require sending your conversations to the cloud. Unlike proprietary chatbots, it gives you control over the model and your data, supporting open tools like Ollama and LM Studio. It's currently in beta, so bugs and incomplete functionality are possible, but you can already try out the scenario of fully offline LLM conversations on Android.
⚡ The gist in 5 seconds - A beta version of Tinariox has been released — a local AI companion for Android .
- It connects to Ollama, LM Studio, or any OpenAI-compatible cloud endpoint.
- It works with voice input, memory, and tasks right on the device.
🔍 What was found A beta APK has been published in the SevernyiVeter/tinariox-app repository.
It's a free local AI companion for Android: it lets you chat with an LLM using either local servers (LM Studio, Ollama) or OpenAI-compatible cloud APIs.
Built-in voice input, memory, and task features are advertised.
How it works: the app follows a local-first paradigm, meaning data and processing stay on the device, while the model is connected via a custom endpoint.
For local use, you can run Ollama or LM Studio on a PC and enter its address in the settings; for the cloud, you can specify any OpenAI-compatible URL.