Mithi v1 Edge

A compact text-to-speech model for 11 Indian languages and 10 voices, built to run on phones, laptops and small devices.

  • Languages: Hindi, English, Bengali, Gujarati, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu.
  • Voices: Satvik, Anirban, Nakul, Vasant, Siddhant (male); Charvi, Ahana, Vaidehi, Malini, Jahnavi (female).
  • Size: 47.6 M parameters; 39 MB of weights (46 MB with the tokenizer files).
  • Weights: compressed (int8 and int4) safetensors.
  • Output: 24 kHz mono audio.
  • License: Apache-2.0, free to use, including commercially. See License.

Try it in your browser: Mithi v1 Edge demo.

License

Mithi is released under the Apache License 2.0 (see LICENSE). It is free to use, for personal, research and commercial projects alike: you can use, modify and redistribute the model and code, and build products on it. Keep the license and the credits (see Acknowledgements) with any copy you share.

Mithi was built entirely on open-source models and tools, each under a permissive license:

Component License
Maya2-native Apache-2.0
Kokoro-82M Apache-2.0
indic-bert MIT
Maya text normaliser Apache-2.0
misaki Apache-2.0

Packages installed with pip keep their own licenses.

Quick start

1. Download the model and install the dependencies (Python 3.10 or newer):

pip install -U huggingface_hub
hf download Oomni/mithi-v1-edge --local-dir mithi-v1-edge
cd mithi-v1-edge
pip install "torch>=2.4" "transformers>=5.0" "safetensors>=0.4" "sentencepiece>=0.2" "numpy>=1.26" "soundfile>=0.12" \
    "misaki[en]>=0.9.4" "espeakng-loader>=0.2.4" "phonemizer-fork>=3.3.2" "spacy>=3.8" "num2words>=0.5"
python -m spacy download en_core_web_sm

espeak-ng comes with the espeakng-loader package, so no system install is needed. A GPU is used if there is one; the model also runs on CPU.

2. Run it from the command line:

python run.py --lang hi --voice ahana --text "आज मौसम कितना अच्छा है! चलो, शाम को कहीं बाहर घूमने चलते हैं।" --out hindi.wav
python run.py --lang en --voice anirban --text "Your appointment is confirmed for Monday at 10:30 AM." --out english.wav

3. Or from Python, inside the mithi-v1-edge folder:

from run import Mithi

tts = Mithi()                                      # loads the model from this folder
audio = tts.synthesize("नमस्ते! आप कैसे हैं?", lang="hi", voice="vaidehi")
tts.save(audio, "namaste.wav")                     # 24 kHz WAV

for voice in tts.voices:                           # all 10 voices
    tts.save(tts.synthesize("வணக்கம்! எப்படி இருக்கீங்க?", lang="ta", voice=voice), f"hello_ta_{voice}.wav")

4. Run on CPU (laptops, Raspberry Pi, phones, servers without a GPU):

python run.py --device cpu --threads 4 --lang hi --voice charvi --text "आज मौसम कितना अच्छा है! चलो, शाम को कहीं बाहर घूमने चलते हैं।" --out hindi_cpu.wav

--threads sets how many CPU cores to use; without it PyTorch uses all of them. From Python:

import torch
torch.set_num_threads(4)

from run import Mithi
tts = Mithi(device="cpu")
tts.save(tts.synthesize("नमस्ते! आप कैसे हैं?", lang="hi", voice="nakul"), "namaste_cpu.wav")

Languages: hi Hindi · en English · bn Bengali · gu Gujarati · kn Kannada · ml Malayalam · mr Marathi · or Odia · pa Punjabi · ta Tamil · te Telugu

Long text is read sentence by sentence automatically.

Voices

Ten voices, five male and five female. Every voice speaks all 11 languages.

Voice --voice Gender
Satvik satvik Male
Anirban anirban Male
Nakul nakul Male
Vasant vasant Male
Siddhant siddhant Male
Charvi charvi Female
Ahana ahana Female
Vaidehi vaidehi Female
Malini malini Female
Jahnavi jahnavi Female

Files

File Contents
run.py everything needed to run the model: command line and Python API
config.json languages, voices and settings
*.safetensors, indic_encoder/ model weights (compressed) and tokenizer
onnx/ files for the browser demo
LICENSE Apache License 2.0

Limitations

  • Expressiveness: intonation and emotion are flatter than in natural human speech.
  • Compact model: it trades some naturalness for size; the larger Mithi models sound slightly more natural.
  • Punjabi with Jahnavi or Vasant may sound less natural than with the other voices.
  • Pronunciation: some words, especially rare names and loanwords, can be mispronounced.
  • Numbers, dates, currency, URLs are read only as well as the text normaliser expands them.
  • Long text is read sentence by sentence; each sentence is spoken without seeing the others.
  • No published evaluation yet: no automated scores and no human listening study (MOS).

Acknowledgements

Downloads last month
66
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Space using Oomni/mithi-v1-edge 1