10月1日 的 Private Mind: Local AI 應用程式分析

Run open AI models like Llama, Gemma, and Qwen on your phone. Offline. Private. Free. No account, no tracking, no subscription. Powered by React Native ExecuTorch. Meet Private Mind, the AI assistant that never phones home. Every conversation runs entirely on your device, so it works in airplane mode, on a plane, or anywhere the internet can't reach. Why Private Mind? · 100% offline: no internet needed, chat anywhere · Total privacy: zero telemetry, your words never leave the phone · Chat with your documents: import PDF or TXT files and get answers with clickable citations, so you can verify every source · Model Hub: curated open models grouped by family, or import your own · Built-in benchmarks: see how fast each model runs on your hardware before you commit · Local speech-to-text: dictate prompts with Whisper, transcribed on-device · Make it yours: custom system prompts, chat search, conversation forking and export Models download once (about 1 to 3 GB each, best over Wi-Fi) and stay on your device. A modern phone with 4 GB+ RAM is recommended for larger models. Not sure? Run the built-in benchmark. Free and open source. Verify the code yourself: github.com/software-mansion-labs/private-mind Built by Software Mansion, core React Native contributors, on React Native ExecuTorch.