text-to-speech (TTS)
EmotiVoice is a multi-voice, prompt-controlled TTS engine that synthesizes emotional speech in English and Chinese, with a web UI and API.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
choose for tokenizer-free multilingual generation if emotion prompts are not required.
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
choose for cloning across 600+ languages; lacks emotion-prompt control.
High-Quality Voice Cloning TTS for 600+ Languages
choose for few-shot voice cloning from minimal data; not for predefined emotion prompts.
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
choose if you already use PaddleSpeech and need a broader speech toolkit.
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.
choose for academic experiments needing many speech tasks; it is heavier.
End-to-End Speech Processing Toolkit
choose for instant cloning without emotion prompts; different API.
Instant voice cloning by MIT and MyShell. Audio foundation model.
choose for cloning with emotional expressiveness, but it lacks EmotiVoice's prompt system.
SoTA open-source TTS
choose for industrial zero-shot TTS with control; not same emotion-prompt interface.
An Industrial-Level Controllable and Efficient Zero-Shot Text-To-Speech System
choose for a complete multilingual TTS stack, but it is heavier than EmotiVoice.
Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
choose for embedded/on-device TTS via ONNX; not a TTS-only service.
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
choose to run Spark-TTS inference, but there's no bundled service layer.
Spark-TTS Inference Code
choose for building a voice agent; for pure TTS it adds unrelated features.
Very low latency speech to text, intent recognition, and text to speech, for building voice agents and interfaces
choose for reproducible audio research; not as a drop-in TTS API.
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.
choose for on-CPU, latency-critical TTS; sacrifices emotion control.
A TTS that fits in your CPU (and pocket)
choose for Apple Silicon native TTS; limited to that ecosystem.
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.
choose for a high-quality multilingual model, but you must build the serving layer.
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.
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
choose for a lightweight local TTS on embedded devices, but it is archived.
A fast, local neural text to speech system
choose for legacy voice cloning demos; it is unmaintained.
Clone a voice in 5 seconds to generate arbitrary speech in real-time
choose as a research toolkit to train custom models, but development has stalled.
🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production
choose for older Chinese voice cloning; it is unmaintained.
🚀Clone a voice in 5 seconds to generate arbitrary speech in real-time
choose for an expressive TTS baseline, but it is stale and unmaintained.
Foundational model for human-like, expressive TTS
choose for zero-shot TTS research, but it is not production-ready.
YourTTS: Towards Zero-Shot Multi-Speaker TTS and Zero-Shot Voice Conversion for everyone
choose to study the classic VITS architecture; too old for production.
VITS: Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech
choose for a lightweight multilingual TTS, but it is not maintained.
High-quality multi-lingual text-to-speech library by MyShell.ai. Support English, Spanish, French, Chinese, Japanese and Korean.
Does the same job, but ships as an app. Useful to a person, not swappable into a codebase.
choose for a free cloud TTS API; it is a wrapper, not a local engine.
Use Microsoft Edge's online text-to-speech service from Python WITHOUT needing Microsoft Edge or Windows or an API key
These projects were analysed and named EmotiVoice among their alternatives. The relationship is not symmetric — how EmotiVoice rates them is a separate judgement, made when EmotiVoice is analysed in its own right.
calls EmotiVoice “Choose it for an actively maintained, pip-installable multi-voice engine with Chinese support and prompt/emotion control.”
Chinese Mandarin tts text-to-speech 中文 (普通话) 语音 合成 , by fastspeech 2 , implemented in pytorch, using waveglow as vocoder, with biaobei and aishell3 datasets
calls EmotiVoice “A prompt-controlled multi-voice TTS engine; pick it for emotion/prompt-based synthesis instead of Coqui's standard multi-speaker models.”
🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production
calls EmotiVoice “For an English/Chinese TTS engine with fine emotion control, if that matters more than ONNX performance.”
Lightning-Fast, On-Device, Multilingual TTS — running natively via ONNX.
calls EmotiVoice “Choose when you need prompt-controlled emotions in a local TTS engine; fewer languages.”
Use Microsoft Edge's online text-to-speech service from Python WITHOUT needing Microsoft Edge or Windows or an API key
calls EmotiVoice “Choose when you need a prompt-controlled multi-voice TTS engine, though it is not updated recently.”
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.
calls EmotiVoice “choose for prompt-controlled emotive TTS from a pretrained engine rather than a trainable multi-model library.”
:robot: :speech_balloon: Deep learning for Text to Speech (Discussion forum: https://discourse.mozilla.org/c/tts)
calls EmotiVoice “choose if you need prompt-controlled emotion in TTS and have GPU.”
A TTS that fits in your CPU (and pocket)
calls EmotiVoice “choose for prompt-controlled emotional multi-voice TTS; code is dormant but usable.”
VITS: Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech
calls EmotiVoice “Choose if you need a prompt-controlled multi-voice engine, but it is stale and unmaintained.”
VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning
calls EmotiVoice “Older emotion-controllable TTS; not updated for years.”
SoTA open-source TTS
calls EmotiVoice “Multi-voice TTS engine, unmaintained; use if you need its prompt-controlled TTS.”
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.
calls EmotiVoice “an unmaintained multi-voice prompt-controlled TTS engine.”
Spark-TTS Inference Code
calls EmotiVoice “Choose when you need a prompt-controlled multi-voice TTS engine and ESPnet's TTS doesn't fit your specific needs.”
End-to-End Speech Processing Toolkit
calls EmotiVoice “Use for prompt-controlled emotional TTS, but no updates in 2 years.”
High-Quality Voice Cloning TTS for 600+ Languages
calls EmotiVoice “Choose when you need emotion-controllable TTS and want a plain PyTorch implementation.”
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.
calls EmotiVoice “Older prompt-controlled TTS engine; use it for emotion control if you accept a two-year-old codebase.”
High-quality multi-lingual text-to-speech library by MyShell.ai. Support English, Spanish, French, Chinese, Japanese and Korean.