parakeet.wgsl – Fast, accurate ASR in the browser, via raw WebGPU & SIMD WASM
Summary
parakeet.wgsl enables fast, accurate NVIDIA Parakeet TDT 0.6B V2 speech transcription entirely in the browser using raw WebGPU compute shaders and SIMD WASM, with a live demo and open-source library.
Similar Articles
@mudler_it: parakeet.cpp now runs NVIDIA Parakeet behind the OpenAI API. Point any OpenAI client at a local server, send an audio, …
parakeet.cpp enables running NVIDIA Parakeet ASR behind the OpenAI API locally with prebuilt Docker images, supporting CPU and CUDA (including arm64) for real-time transcription with word timestamps.
@badlogicgames: what a wonderful project: parakeet.cpp https://github.com/mudler/parakeet.cpp… GGML based parakeet inference pipeline t…
parakeet.cpp is a fast, dependency-light C++17 inference pipeline for NVIDIA's NeMo Parakeet speech recognition models, built on ggml. It achieves byte-identical transcripts to NeMo with significant speedups on CPU and GPU.
I fine-tuned Parakeet 0.6B for medical ASR — open weights, local Mac/CUDA/CPU
Omi Health founder fine-tuned NVIDIA's Parakeet TDT 0.6B for medical ASR, releasing open-weights model Omi Med STT v1 that achieves competitive medical-WER while running locally on Mac, CUDA, or CPU.
I ported NVIDIA Parakeet (speech-to-text) to ggml: same output as NeMo, faster, GGUF-quantized, no Python
NVIDIA's Parakeet speech-to-text models have been ported to pure C++/ggml, achieving byte-identical output to NeMo, up to 5x faster inference on GPU, and quantized GGUF variants for efficient deployment anywhere without Python or PyTorch.
Parrot Speech-to-text API
Parrot Speech-to-text API offers fast and accurate transcription for production-grade voice agents.