Created by Neuphonic - building faster, smaller, on-device voice AI
State-of-the-art Voice AI has been locked behind web APIs for too long. NeuTTS Air is the world’s first super-realistic, on-device, TTS speech language model with instant voice cloning. Built off a 0.5B LLM backbone, NeuTTS Air brings natural-sounding speech, real-time performance, built-in security and speaker cloning to your local device - unlocking a new category of embedded voice agents, assistants, toys, and compliance-safe apps.
Key Features
🗣Best-in-class realism for its size - produces natural, ultra-realistic voices that sound human
📱Optimised for on-device deployment - provided in GGML format, ready to run on phones, laptops, or even Raspberry Pis
👫Instant voice cloning - create your own speaker with as little as 3 seconds of audio
🚄Simple LM + codec architecture built off a 0.5B backbone - the sweet spot between speed, size, and quality for real-world applications
[!CAUTION]
Websites like neutts.com are popping up and they're not affliated with Neuphonic, our github or this repo.
We are on neuphonic.com only. Please be careful out there! 🙏
Model Details
NeuTTS Air is built off Qwen 0.5B - a lightweight yet capable language model optimised for text understanding and generation - as well as a powerful combination of technologies designed for efficiency and quality:
Audio Codec: NeuCodec - our proprietary neural audio codec that achieves exceptional audio quality at low bitrates using a single codebook
Format: Available in GGML format for efficient on-device inference
Responsibility: Watermarked outputs
Inference Speed: Real-time generation on mid-range devices
Power Consumption: Optimised for mobile and embedded devices
The requirements file includes the dependencies needed to run the model with PyTorch. When using an ONNX decoder or a GGML model, some dependencies (such as PyTorch) are no longer required.
The inference is compatible and tested on python>=3.11.
pip install -r requirements.txt
Basic Example
Run the basic example script to synthesize speech:
bash
1python -m examples.basic_example \2 --input_text "My name is Dave, and um, I'm from London"\3 --ref_audio samples/dave.wav \4 --ref_text samples/dave.txt
5
To specify a particular model repo for the backbone or codec, add the --backbone argument. Available backbones are listed in NeuTTS-Air huggingface collection.
Several examples are available, including a Jupyter notebook in the examples folder.
Simple One-Code Block Usage
python
1from neuttsair.neutts import NeuTTSAir
2import soundfile as sf
34tts = NeuTTSAir( backbone_repo="neuphonic/neutts-air-q4-gguf", backbone_device="cpu", codec_repo="neuphonic/neucodec", codec_device="cpu")5input_text ="My name is Dave, and um, I'm from London."67ref_text ="samples/dave.txt"8ref_audio_path ="samples/dave.wav"910ref_text =open(ref_text,"r").read().strip()11ref_codes = tts.encode_reference(ref_audio_path)1213wav = tts.infer(input_text, ref_codes, ref_text)14sf.write("test.wav", wav,24000)15
Tips
NeuTTS Air requires two inputs:
A reference audio sample (.wav file)
A text string
The model then synthesises the text as speech in the style of the reference audio. This is what enables NeuTTS Air’s instant voice cloning capability.
Example Reference Files
You can find some ready-to-use samples in the examples folder:
samples/dave.wav
samples/jo.wav
Guidelines for Best Results
For optimal performance, reference audio samples should be:
Mono channel
16-44 kHz sample rate
3–15 seconds in length
Saved as a .wav file
Clean — minimal to no background noise
Natural, continuous speech — like a monologue or conversation, with few pauses, so the model can capture tone effectively
