debpalash / VoiceStudio

VoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice design, video dubbing, dictation, transcription & audiobook creation in 646 languages.

C4rohan / Lampline

Self-hosted voice agent platform: import your Dialogflow CX and CX Agent Studio agents, build and test them, and take calls on your own hardware. 20 model providers, phone and SIP through jambonz.

ahmedbadr-dev / voice-ai-assistant

Python voice AI assistant: microphone STT, Google Gemini answers, offline TTS with pyttsx3. Clean env-based setup.

jamiepine / voicebox

The open-source AI voice studio. Clone, dictate, create.

GetStream / Vision-Agents

Open Vision Agents by Stream. Build voice and vision agents quickly with any model or video provider. Uses Stream's edge network for ultra-low latency.

dograh-hq / dograh

Open source voice AI platform. Self-hosted alternative to Vapi and Retell. On Prem, BYOK across Speech to Speech or LLM/STT/TTS, with a visual workflow builder, MCP native and telephony support.

nicedreamzapp / claude-code-local

Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server. 6 fighters incl. Muse-Glimmer 30B (now multimodal — reads images, abliterated), Gemma 4 31B, Qwen 3.5 122B (65 tok/s), DeepSeek V4 Flash (1M ctx). Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows.

PatterAI / Patter

Open-source voice-AI SDK. The Vapi/Retell alternative for builders who want to own the stack. Give your AI agent a phone number in 4 lines — Python and TypeScript, MIT licensed, Twilio, Telnyx, and Plivo.

TheStageAI / TheWhisper

Optimized Whisper models for streaming and on-device use

bolna-ai / bolna

Conversational voice AI agents

usemoss / moss

The retrieval layer for production AI systems. Lightning-fast (<10ms) search without vector databases. Built for browser, edge, on-device, and cloud.

fluxions-ai / vui

Vui Nano — a small, context-aware text-to-speech model trained on real conversations. 219M active params (305M total), Apache 2.0, voice cloning, streaming, runs on CPU (dependency-free C build). Ships with a full real-time voice assistant: WebRTC, ASR, local LLM, OpenAI Realtime API compatible.

Johnson-Durui / Companion-Space

本地优先的二次元陪伴学习应用 · Local-first anime companion and study app.

allmodels-io / hermes-speech

Hermes Agent TTS and STT plugin for AllModels speech models and voices.

aws-samples / sample-voice-ai-agent-studio

—

TheSmokeDev / hermes-talk

Hermes Talk: realtime duplex voice for Hermes Agent. Rides your ChatGPT or X Premium subscription, no API key needed. OpenAI, Grok, Gemini. Terminal, Discord, dashboard.

bluejay-ai-dev / mivas-bench

Multi-Industry Voice Agent Simulation Bench (MIVAS) is an indicator of voice AI performance across economic sectors. MIVAS is comprised of tasks and verifiers with granular rewards executed in production-grade multi-agent RL environments across industries where voice AI is being adopted the fastest.

rehan-remade / open-annie

A voice-driven 3D character in the browser. The model that talks is not the model that moves: GPT-Live-1 talks, Jev moves her, three.js + VRM render her, a thin broker runs on fal.

arbazsiddiqui / voxlock

The best performing voice isolator there is. Keeps the caller's voice and removes the other voices in the room before speech-to-text hears them. Open source, streaming, runs on CPU and in the browser. pip install voxlock

DINAKAR-S / voice-agent-starter-kit

Self-hosted multilingual (English + 5 Indic) phone voice agent template - LiveKit Agents + Sarvam STT/TTS + OpenAI + Vobiz SIP. ~700ms-1.2s latency, under Rs 5.5/min. Demo: a real-estate site-visit booking agent.

Mayankpratapsingh022 / Turn-Detection-Model-Whisper-Tiny-Dual-Scale-Attention-Classification-Head-for-Low-Latency-Voice-AI

Low-latency, audio-native turn detection model using Whisper Tiny with a dual-scale attention classification head. Optimized for Hindi, Hinglish and Indian English, including fillers, incomplete thoughts and natural pauses. Includes causal training, hard-negative mining, calibrated endpointing.