text-to-speech (TTS)
MeloTTS is a high-quality multi-lingual text-to-speech library supporting English, Spanish, French, Chinese, Japanese, and Korean.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
Focuses on instant voice cloning rather than a full multi-speaker library; choose it when cloning a specific voice is the priority.
Instant voice cloning by MIT and MyShell. Audio foundation model.
Tokenizer-free generative TTS for multilingual and creative voice design; use it for a modern generative model, not a lightweight library.
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
State-of-the-art open-source TTS with expressive output; pick it if you prioritize latest quality and can handle higher compute.
SoTA open-source TTS
Industrial zero-shot TTS with controllable prosody; use it for production-grade control rather than a simple multi-lingual drop-in.
An Industrial-Level Controllable and Efficient Zero-Shot Text-To-Speech System
Large multilingual voice generation model with full-stack ability; choose it for extensive language coverage and training options.
Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
Embedded-friendly runtime with TTS and ASR via ONNX; use it for offline/low-resource deployment without internet.
Speech-to-text, text-to-speech, speaker diarization, speech enhancement, source separation, and VAD using next-gen Kaldi with onnxruntime without Internet connection. Support embedded systems, Android, iOS, HarmonyOS, Raspberry Pi, RISC-V, RK NPU, Axera NPU, Ascend NPU, x86_64 servers, websocket server/client, support 12 programming languages
Inference code for a recent TTS model; choose it if you specifically want Spark-TTS's voice cloning and language features.
Spark-TTS Inference Code
Voice cloning TTS covering 600+ languages; pick it for extreme language coverage over MeloTTS's six languages.
High-Quality Voice Cloning TTS for 600+ Languages
Ultra-lightweight CPU-only TTS; choose it when running on a laptop or edge device with minimal resources.
A TTS that fits in your CPU (and pocket)
Apple Silicon (MLX) TTS/STT library; select it for tight Apple ecosystem integration rather than cross-platform Python.
A text-to-speech (TTS), speech-to-text (STT) and speech-to-speech (STS) library built on Apple's MLX framework, providing efficient speech analysis on Apple Silicon.
High-quality open-weight multilingual TTS model; use it when audio quality matters more than library lightness.
Zonos-v0.1 is a leading open-weight text-to-speech model trained on more than 200k hours of varied multilingual speech, delivering expressiveness and quality on par with—or even surpassing—top TTS providers.
Few-shot voice cloning from 1 minute of audio; ideal for personalized cloning rather than broad multi-lingual coverage.
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
Comprehensive speech toolkit with TTS, ASR, and more; choose it if you want a full speech pipeline in PaddlePaddle.
Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.
Thin Python client for Microsoft Edge online TTS; use it when you want cloud voices without running a local model.
Use Microsoft Edge's online text-to-speech service from Python WITHOUT needing Microsoft Edge or Windows or an API key
Research toolkit for audio/speech generation, includes TTS models; pick it for reproducible experiments rather than a simple library.
Amphion (/æmˈfaɪən/) is a toolkit for Audio, Music, and Speech Generation. Its purpose is to support reproducible research and help junior researchers and engineers get started in the field of audio, music, and speech generation research and development.
End-to-end speech processing toolkit with many TTS recipes; choose it for a research framework with broad support.
End-to-End Speech Processing Toolkit
TTS built by inverting Whisper, a different architecture; use it if you want that specific approach and unique voice.
An Open Source text-to-speech system built by inverting Whisper.
100M-parameter multilingual TTS for CPU and voice cloning; choose it for real-time CPU inference with cloning.
MOSS-TTS-Nano is an open-source multilingual tiny speech generation model from MOSI.AI and the OpenMOSS team. With only 0.1B parameters, it is designed for realtime speech generation, can run directly on CPU without a GPU, and keeps the deployment stack simple enough for local demos, web serving, and lightweight product integration.
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
A mature TTS toolkit with a large model zoo, but unmaintained for two years; choose it if you need many models and can accept stale releases.
🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production
Fast local neural TTS, archived and last touched over a year ago; still useful for small ONNX models but no longer maintained.
A fast, local neural text to speech system
Older prompt-controlled TTS engine; use it for emotion control if you accept a two-year-old codebase.
EmotiVoice 😊: a Multi-Voice and Prompt-Controlled TTS Engine
Older fast TTS model supporting many languages; useful for a tiny model in legacy projects, but stale.
A fast Text-to-Speech (TTS) model. Work well for English, Mandarin/Chinese, Japanese, Korean, Russian and Tibetan (so far). 快速语音合成模型，适用于英语、普通话/中文、日语、韩语、俄语和藏语（当前已测试）。
Archived Tacotron 2 implementation for multilingual research; use only for academic baseline comparisons.
An implementation of Tacotron 2 that supports multilingual experiments with parameter-sharing, code-switching, and voice cloning.
These projects were analysed and named MeloTTS among their alternatives. The relationship is not symmetric — how MeloTTS rates them is a separate judgement, made when MeloTTS is analysed in its own right.
calls MeloTTS “For a Python-native TTS with high-quality multilingual voices, if you don't need ONNX cross-platform deployment.”
Lightning-Fast, On-Device, Multilingual TTS — running natively via ONNX.
calls MeloTTS “Choose when you need a lightweight multilingual local TTS library; good for low-resource deployment.”
Use Microsoft Edge's online text-to-speech service from Python WITHOUT needing Microsoft Edge or Windows or an API key
calls MeloTTS “Older multilingual TTS library with limited maintenance.”
SoTA open-source TTS
calls MeloTTS “A lighter multilingual TTS without zero-shot voice cloning; choose for low-resource or highly efficient deployment.”
Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
calls MeloTTS “choose if you need a lightweight multi-lingual TTS and accept an unmaintained project.”
A TTS that fits in your CPU (and pocket)
calls MeloTTS “choose for a lightweight multilingual TTS, but it is not maintained.”
EmotiVoice 😊: a Multi-Voice and Prompt-Controlled TTS Engine