{"repo":"m15-ai/Local-Voice","free":true,"listed":false,"github":"https://github.com/m15-ai/Local-Voice","clone":"git clone https://github.com/m15-ai/Local-Voice.git","description":"A real-time, offline voice assistant for Linux and Raspberry Pi. Uses local LLMs (via Ollama), speech-to-text (Vosk), and text-to-speech (Piper) for fast, wake-free voice interaction. No cloud. No APIs. Just Python, a mic, and your voice.","language":"Python","stars":22,"topics":["llm","offline-ai","ollama","piper","python","raspberry-pi","speech-recognition","stt","text-to-speech","tts"],"license":"MIT","category":"ai-agents","readme_excerpt":"Local Voice Assistant (Offline, Real-Time AI) Lightweight, low-latency voice assistant running fully offline on a Raspberry Pi or Linux machine. Powered by PyAudio, Vosk STT, Piper TTS, and local LLMs via Ollama. --- 🎯 Features - 🎙️ Microphone Input using PyAudio - 🔊 Real-Time Transcription with Vosk - 🧠 LLM-Powered Responses using Ollama with models like gemma2:2b , qwen2.5:0.5b - 🗣️ Natural Voice Output via Piper TTS - 🎛️ Optional Noise & Filter FX using SoX for realism - 🔧 ALSA Volume Control - 🧩 Modular Python code ready for customization --- 🛠 Requirements - Raspberry Pi 5 or Linux desktop - Python 3.9+ - PyAudio, NumPy, requests, soxr, pydub, vosk - SoX + ALSA utilities - Ollama with one or more small LLMs (e.g., Gemma or Qwen) - Piper TTS with ONNX voice models Install dependencies: ⚙️ JSON Configuration Place a config file at va config.json: Note: if the configuration file is not found, defaults withing the main python app will be used: 🔁 What history length Means The history length setting controls how many previous exchanges (user + assistant messages) are included when generating each new reply. - A value of 6 means the model receives the last 6 exchanges, plus the system prompt. - This allows the assistant to maintain short-term memory for more coherent conversations. - Setting it lower (e.g., 2 ) increases speed and memory efficiency. ✅ requirements.txt If you plan to run this on a Raspberry Pi, you may also need: 🐍 Install with Virtual Environment 💡 ","default_branch":null,"files":null,"tree":[],"storefront":"/r/m15-ai","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/m15-ai/Local-Voice/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}