voice-assistant

18 repositories

agentscope-ai/agentscope

Agent framework built around transparency and developer trust — AgentScope is a Python framework from Alibaba for building multi-agent, multi-modal AI systems, including ReAct, voice, and realtime agents with MCP support. Backed by an academic paper, it focuses on making agent behavior visible and debuggable rather than treating them as black boxes. Useful for developers who need to inspect and trust complex agent workflows.

Python🇨🇳 Greater ChinaAttention 39
★36,320
Fosowl/agenticSeek

A fully local, voice-enabled agent that browses and codes for you — AgenticSeek positions itself as a private, local alternative to Manus AI: it browses the web, writes and debugs code, and plans multi-step tasks using local reasoning models like DeepSeek-R1, with no cloud or subscription fees. Voice interaction is in progress. Useful for privacy-conscious developers wanting autonomous agents without API costs.

Python🇪🇺 EuropeAttention 20
★27,676
pipecat-ai/pipecat

Open Source framework for voice agents, multimodal apps, and realtime AI. Maintained by Daily and the community.

PythonAttention 37
★17,351
isair/jarvis

A voice-first AI companion that stays entirely on your machine — Jarvis is a desktop assistant that listens for its name mid-conversation, keeps a rolling transcript, and replies with voice or text using local models via Ollama or OpenAI-compatible servers. It redacts sensitive data before saving to its local diary and knowledge graph, appealing to privacy-conscious users wanting a subscription-free alternative to cloud assistants.

Python🇪🇺 EuropeAttention 65
★2,381
JosssphZhou/poki

Give AI agents hands-on control of a real iPhone via MCP — Poki lets AI agents operate a genuine iPhone through a local MCP server and agent, with no jailbreak or extra hardware required. It plugs into existing MCP clients like Claude Code or Codex, letting you automate real iOS interactions using your own subscription instead of a separate model account.

SwiftAttention 86
★1,668
wilbowes/EchoMuse

Repurpose a 2016 Echo Dot as a cloud-free HA voice satellite — EchoMuse replaces the firmware on a second-gen Echo Dot so it works as a local ESPHome voice satellite for Home Assistant, using Assist instead of Alexa. No Amazon account or cloud needed; it runs as an add-on or Docker container. Useful for hobbyists wanting to repurpose cheap secondhand hardware, though setup requires a USB unlock step (firebreak) with some brick risk.

PythonAttention 55
★1,522
SaarthurR/atticus-mac

A voice assistant that runs your Mac with Claude Pro — Atticus is a free, open-source macOS voice assistant triggered by a hotkey. It uses your existing Claude Pro or Max plan instead of an API key, skips AI for repeated actions (0.08s app launches), and learns your shortcuts and personal context over time.

Python🇺🇸 North America
★801
HuskerMinion/techo5

Replace Amazon's firmware on the Echo Show 5 with Linux and Home Assistant — TECHO5 flashes the Echo Show 5 (1st/2nd gen) with Alpine Linux, turning it into a local Home Assistant voice satellite via the ESPHome API, with on-device wake word, touch screen support, SSH access and A/B OTA updates. No Android, no Alexa, no Amazon cloud. A sibling project (ATHENA) extends this to the Echo Show 8 2nd gen.

GoAttention 64
★765
ygelfand/echolocal

Repurposing Echo hardware as offline Home Assistant satellites — EchoLocal is a pure-Go firmware replacement that strips Amazon's cloud services from Echo Dot and Echo Show devices, turning them into local voice satellites, media players, and displays. It speaks ESPHome's native API, so Home Assistant auto-discovers these repurposed devices like any other ESPHome hardware, appealing to privacy-focused smart home tinkerers.

GoAttention 56
★602
adventists-ai/sophon-realtime

Open full-duplex voice models that hear your tone and speak with one. One GPU with vLLM-Omni, or fully offline on a Mac with MLX. Zero-shot voice cloning, OpenAI Realtime / GPT-Live compatible, free hosted API trial. Apache-2.0.

Python🇨🇳 Greater ChinaAttention 86
★482
JMS1717/ming-voice

Low-latency local speech-to-speech voice assistant with Ming-3.0-audio-tiny on AMD and Nvidia GPUs. Easy setup on Windows and Linux

PythonAttention 40
★450
Dodaios/Dodaios

A chief-of-staff app for running nine coding agents side by side — Dodaios (formerly Personal Jarvis) is an open-source desktop workspace that runs Claude Code, Codex and seven other coding agents together, coordinated by one assistant with voice input. It uses your own plans and API keys, runs locally, and supports MCP for agent orchestration.

PythonAttention 68
★438
raullenchai/esp32-thinks

A multi-model Muse: personal AI voice assistant on ESP32 - wake word or knob, any OpenAI-compatible brain (STT/chat/TTS mix-and-match), streaming replies, out of the box. Built on muse-gadget-sdk.

Shell🇺🇸 North AmericaAttention 0
★256
CapriChooms/CapriCore

Capri — an AI desktop companion for Windows with local/API models, Hugging Face, optional voice, memory and permission-based desktop tools.

JavaScript
★250
vahmoh/Veda

A Rust operating system with a built-in voice agent — Veda is a Rust-built OS, from UEFI bootloader to microkernel, where a voice agent controls apps through typed actions instead of screen scraping. It asks for consent before destructive actions, runs Linux drivers isolated in a VM, and ships a full desktop with browser, terminal and Wi-Fi.

Rust🇪🇺 Europe
★109
ebongard/renfield

Self-hosted AI assistant with voice satellites, 100+ tool integrations, RAG knowledge base, and full offline capability. Built with FastAPI + React + Ollama.

PythonAttention 39
★47
bomby-app/bomby

Talk to your coding agent. A Mac app that hands what you say to Claude Code, Codex, OpenCode, goose or omp, and tells you when it's done.

TypeScript
★23
scalekit-developers/livekit-scalekit-voice-agent

A voice agent prototype that keeps OAuth tokens away from the LLM — This is a prototype combining LiveKit Agents with Scalekit AgentKit so a voice assistant can call tools like Google Calendar on a user's behalf without exposing raw OAuth tokens to the LLM. It targets developers exploring secure, per-user authenticated tool calling for voice AI; note it's still unvalidated against live credentials.

TypeScript
★0