MeloTTS

MeloTTS TTS

MyShell.ai's multilingual model that runs near real-time on CPU and supports multiple English accents.

Sign up for 5,000 character limit

Wrap your text in SSML tags for precise control:

<speak><prosody rate="slow">Slow speech</prosody></speak>

Tags the selected model understands — click to drop one into your text where it happens:

This model reads plain text, so inline tags are ignored. For tag-based emotion, switch to an expressive model like Orpheus or Bark.

Define custom pronunciations (word = pronunciation):

-12 +12
0.5x 2.0x
Free with Piper, VITS, MeloTTS
Your generated audio will appear here. Choose a model, enter text, and click Generate.
Audio Generated Successfully
0:00
Download Audio Download .srt Link expires in 24h
Free tier: personal use. Commercial license from $5/mo
Love TTS.ai? Tell your friends!

About MeloTTS

MeloTTS by MyShell.ai is a production-oriented multilingual library built on a VITS-based, multi-accent architecture. It is engineered for speed, processing text at near real-time on CPU alone, with optional GPU inference for extra throughput. A distinguishing feature is its accent range in English — American, British, Indian, and Australian — alongside support for Spanish, French, Chinese, Japanese, and Korean. With a 5,000-character cap and low latency, it's a dependable choice for production apps that need fast, multilingual, accent-aware speech without heavy hardware.

Best for: Production applications needing fast, multilingual TTS

Browse all MeloTTS voices

At a glance

Developer
MyShell.ai
License
MIT
Tier
free
Speed
fast
Voice cloning
No
Languages
English, Spanish, French, Chinese, Japanese, Korean
Max characters
5000

MeloTTS voices

Chinese

Chinese
Free Female

English British

English
Free Female

English US

English
Free Female

French

French
Free Female

Japanese

Japanese
Free Female

Korean

Korean
Free Female

Spanish

Spanish
Free Female

MeloTTS TTS — FAQ

Yes. MeloTTS is CPU-optimized and runs at near real-time speed on CPU alone, with GPU inference optional. It needs only about 0.5GB VRAM if a GPU is used.

MeloTTS offers American, British, Indian, and Australian English accents, plus Spanish, French, Chinese, Japanese, and Korean — 6 languages in total.

Yes. MeloTTS is MIT-licensed and in the free tier. It does not support voice cloning.
← All voices