# LLMVoX

> LLMVoX: Autoregressive Streaming Text-to-Speech Model for Any LLM

## Facts
- **Category:** [Voice Agents](https://agentsearchengine.app/category/voice-agents)
- **Type:** SDK / library
- **Pricing:** Open source
- **GitHub stars:** 308
- **Language:** Python
- **Last repo activity:** 1 year ago (May 16, 2025)
- **Official site:** https://mbzuai-oryx.github.io/LLMVoX/
- **Source code:** https://github.com/mbzuai-oryx/LLMVoX
- **Listing on Agent Search Engine:** https://agentsearchengine.app/agents/llmvox
- **Data verified:** July 2026

## About LLMVoX
Sambal Shikhar, Mohammed Irfan K, Sahal Shaji Mullappilly, Fahad Khan, Jean Lahoud, Rao Muhammad Anwer, Salman Khan, Hisham Cholakkal LLMVoX is a lightweight 30M-parameter, LLM-agnostic, autoregressive streaming Text-to-Speech (TTS) system designed to convert text outputs from Large Language Models into high-fidelity streaming speech with low latency. Our approach achieves significantly lower Word Error Rate compared to speech-enabled LLMs while operating at comparable latency and speech quality. Key features: - Lightweight & Fast: Only 30M parameters, delivering speech with end-to-end latency as low as 300ms - LLM-Agnostic: Just plug with any existing LLM and Vision-Language Models without requiring fine-tuning or architectural modifications. - Multi-Queue Streaming: Enables continuous, low-latency speech generation and infinite-length dialogues - Multilingual Support: Easily adaptable to new languages with only dataset adaptation

_From the project's README._

## Install

```
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
```

## Alternatives
LLMVoX sits in voice agents. Related tools indexed here:
- [xiaozhi-esp32-server](https://agentsearchengine.app/agents/xiaozhi-esp32-server) — 本项目为xiaozhi-esp32提供后端服务，帮助您快速搭建ESP32设备控制服务器。Backend service for xiaozhi-esp32, helps you quickly build an ESP32 device control server. · Open source · Infrastructure · 10k★ · JavaScript · MIT · active 5 days ago
- [ten-vad](https://agentsearchengine.app/agents/ten-vad) — Voice Activity Detector (VAD): low-latency, high-performance and lightweight · Open source · SDK / library · 2.2k★ · C · active 5 months ago
- [bailing](https://agentsearchengine.app/agents/bailing) — 百聆 是一个类似GPT-4o的语音对话机器人，通过ASR+LLM+TTS实现，集成DeepSeek R1等优秀大模型，接入openClaw，真正的个人语音助手，时延低至800ms，Mac等低配置也可运行，支持打断 · Open source · Agent · 1.7k★ · Python · MIT · active 3 months ago
- [RCLI](https://agentsearchengine.app/agents/rcli) — Talk to your Mac, query your docs, no cloud required. On-device voice AI + RAG · Open source · Agent · 1.5k★ · C++ · MIT · active 4 months ago
- [CyberVerse](https://agentsearchengine.app/agents/cyberverse) — Self hosted, real-time digital human agent platform. Build voice-first AI agents with WebRTC, persona memory, tools, RAG, and optional digital-human video. · Open source · Platform · 1.5k★ · Python · GPL-3.0 · active 4 days ago
- [Patter](https://agentsearchengine.app/agents/patter) — Open-source voice-AI SDK. The Vapi/Retell alternative for builders who want to own the stack. · Open source · SDK / library · 1.0k★ · Python · MIT · active 1 day ago
- [rustpbx](https://agentsearchengine.app/agents/rustpbx) — A PBX written by rust · Open source · Infrastructure · 736★ · Rust · MIT · active today
- [echokit_server](https://agentsearchengine.app/agents/echokit-server) — Open Source Voice Agent Platform · Open source · Platform · 589★ · Rust · GPL-3.0 · active 5 months ago

---
Source: [Agent Search Engine](https://agentsearchengine.app) — an independent, hand-curated index of AI agents, MCP servers, and frameworks. Rankings are never sold; sponsored placements are labeled and never numbered. Full directory: https://agentsearchengine.app/llms-full.txt