Voice Assistant Python Packages

pipecat-ai

Open Source framework for voice and multimodal conversational AI

ovos-spec-tools

Reference implementation of the OVOS formal specifications — sentence template expander, locale resource loader, dialog renderer, language matching, and a locale linter

dv-pipecat-ai

Open Source framework for voice and multimodal conversational AI

hivemind-bus-client

Python client library for HiveMind — encrypted, routed bus client with WebSocket, HTTP, async, and MQTT transports (the library for building satellites)

achatbot

An open source chat bot architecture for voice/vision (and multimodal) assistants, local(CPU/GPU bound) and remote(I/O bound) to run.

hivemind-core

Central HiveMind hub server: connects satellite devices and AI agents to an OVOS instance over an encrypted mesh protocol

superdialog

Low latency framework for realtime conversations.

nanowakeword

A lightweight, wake word detection engine. Train custom, high-accuracy models with minimal effort.

penglai

蓬莱 Penglai — 住在你飞书和微信里的中文 AI 管家 · Your personal AI butler in Feishu & WeChat (powered by GenericAgent, kernel untouched)

easyspeak-linux

Voice control for Linux desktops. Fully local, no cloud, Wayland-native.

stark-engine

S.T.A.R.K. - Speech And Text Algorithmic Recognition Kit

pvrhino

On-device Speech-to-Intent engine powered by deep learning

hivemind-mic-satellite

Lightweight HiveMind microphone satellite — streams audio to the hive for server-side STT/TTS (ideal for low-power devices)

hivemind-voice-relay

HiveMind voice relay — local wake word, remote STT/TTS via the hivemind-audio-binary-protocol plugin

hivemind-voice-sat

OVOS voice satellite for HiveMind — full on-device voice stack (microphone, VAD, wake word, STT, TTS)

omnilimb-face

hermes-agent plugin: a VTuber avatar (Live2D/Live3D) with hands-free voice (VAD+STT), real-time barge-in, lip-sync and expressions. Open-LLM-VTuber /client-ws compatible. Part of the omnilimb family.

voxpulse

A lightweight CNN-based framework to train custom wake words offline. Features automated data augmentation, dynamic pathing, and RMS silence gating

attenlabs-saa

Addressee detection for voice agents: device-directed speech detection that runs before STT, so background speech, side conversations, and the agent's own TTS echo never trigger it. No wake word, model-agnostic, drop-in for LiveKit, Pipecat, ElevenLabs, Twilio, and OpenAI. The layer your VAD and turn detection are missing.

iflow-mcp-astrabert-tysva

Learn TypeScript chatting effortlessly with AI

voip

Python VoIP library of the AI age

saa-livekit-client

Addressee detection for voice agents: device-directed speech detection that runs before STT, so background speech, side conversations, and the agent's own TTS echo never trigger it. No wake word, model-agnostic, drop-in for LiveKit, Pipecat, ElevenLabs, Twilio, and OpenAI. The layer your VAD and turn detection are missing.

saa-pipecat-client

Addressee detection for voice agents: device-directed speech detection that runs before STT, so background speech, side conversations, and the agent's own TTS echo never trigger it. No wake word, model-agnostic, drop-in for LiveKit, Pipecat, ElevenLabs, Twilio, and OpenAI. The layer your VAD and turn detection are missing.

violawake

Open-source wake word SDK — production TemporalCNN model, 8-phase training pipeline, ONNX inference, Python-first. The open alternative to Porcupine.

leva-tts

A production-oriented Levantine Text-to-Speech pipeline built on a fine-tuned XTTS-v2 optimized for real-time conversational agents.