Neuphonic just integrated your TTS in my Cloudflare-hosted Voice AI and the voices are very good. 👍

i literally want to continue finetune xcodec2 to upscale from 16k to 24k, but found out huggingface.co/neuphonic/neuc… multilingual 50TPS single codebook 24k sample rate, banger Neuphonic I also tested on malaysian context, super banger compared to other speech tokenizers.

i will open source continue pretraining qwen on emilia yodas voice conversion using neucodec, thanks Neuphonic !

Neuphonic Open-Sources NeuTTS Air: A 748M-Parameter On-Device Speech Language Model with Instant Voice Cloning Neuphonic’s NeuTTS Air is an open-source, ~0.7B-parameter text-to-speech speech LM designed for real-time, on-device CPU inference, distributed in GGUF quantizations

neuphonic tts model is so goated, it is almost at par with the quality of el labs. guess i might be able to run it locally and use it for pod.

on-device text-to-speech with instant voice cloning from just a few seconds of audio

Just dropped on HF — NeuTTS Air Next-gen on-device TTS that matches cloud-level quality while staying fully open source. > Real-time speech synthesis on CPU/GPU > 3-second voice cloning, no cloud or data upload > Compact: under 200 MB, runs on mobile and edge devices >

🙏 🙌🏽

NeuTTS Air: A free TTS model that lets you clone voices on your own devices. 👉 scriptbyai.com/on-device-voic… 🗣️ Best-in-class realism for its size 📱 Optimized for on-device deployment 👫 Instant voice cloning with 3 seconds of audio 🚄 Simple and fast architecture

10. NeuTTS launched NeuTTS Air, a next-gen open-source on-device TTS model. It offers real-time synthesis, 3-second voice cloning, and cloud-level quality in a compact design.

Refreshing to see Neuphonic, a London-based seed startup that raised just a few million, top the most trending models on Hugging Face today. They manage to stand out amongst 2M public models & giant corporations from the US and China. Good example that everyone can

NeuTTS Air: super-realistic TTS model with instant voice cloning it brings natural speech to phones, laptops, even Raspberry Pis Clone any voice with just 3 seconds of audio replicate.com/lucataco/neutt…

For reference, codecs near this size use brute force scaling. Post with technical details: zdtech.substack.com/p/grouchy-my-a… More audio samples: zdisket.github.io/grouchysite/ Yet again, thanks Hot Aisle and AI at AMD for compute. PyTorch/ROCm is very stable and withstands all my experimenting.

For years, creating lifelike AI voices meant depending on big clouds, expensive GPUs, and proprietary APIs. That era is over. Meet NeuTTS, built by Neuphonic, where studio-grade realism, instant cloning, and real-time speech generation all run locally on your device.

@SpheronFDN Neuphonic This is amazing, just tried the demo. The voice is coming super realistic.

Are we the only ones sensing a paradigm shift in AI? 👀 We’re clearly transitioning from an era dominated by massive, closed LLM APIs to one where teams train, fine-tune, and run their own open-source models - smaller, faster, and more specialized. Just look at the past few

Meet VieNeu-TTS, the first-ever Vietnamese Text-to-Speech model that runs entirely on your personal device. Built on the Qwen 0.5B LLM and fine-tuned from NeuTTS Air, it delivers hyper-realistic, natural voices with real-time inference, no heavy hardware dependency.