NeuTTS Nano: 120M TTS Model Runs on Raspberry Pi
NeuTTS Nano is a compact 120-million parameter text-to-speech model optimized to run efficiently on resource-constrained devices like Raspberry Pi, delivering
Someone found a tiny text-to-speech model that actually runs on Raspberry Pis and similar hardware.
NeuTTS Nano is only 120M parameters (3x smaller than its predecessor) and comes in GGML format, which means it works on stuff like Jetson boards and mobile devices without melting them.
The interesting part: it can clone voices from just 3 seconds of audio while fitting in tight RAM constraints.
Grab it here:
- GitHub: https://github.com/neuphonic/neutts
- HuggingFace: https://huggingface.co/neuphonic/neutts-nano
- Live demo: https://huggingface.co/spaces/neuphonic/neutts-nano
Pretty useful if you’re building smart home devices or robotics projects where every megabyte counts. The GGML format makes deployment way simpler than dealing with typical PyTorch models on embedded systems.
The team’s curious about Real-Time Factor benchmarks on different hardware, so people have been testing it on everything from Raspberry Pi 4s to phone CPUs.
Related Tips
Nvidia's DMS Cuts LLM Memory Usage by 8x
Nvidia introduces Dynamic Memory Scheduling that reduces large language model memory consumption by eight times, enabling more efficient AI inference and
Unsloth Kernels: 12x Faster MoE Training, 12GB VRAM
Unsloth Kernels achieves 12x faster Mixture of Experts model training while using only 12GB of VRAM through optimized kernel implementations and memory
Unsloth Kernels: Fine-Tune 30B MoE on Consumer GPUs
Unsloth Kernels enables efficient fine-tuning of 30 billion parameter Mixture of Experts models on consumer-grade GPUs through optimized memory management and