What's the lowest-latency fully local voice loop you've built?
A typical cascade is voice activity detection, a turn detector, streaming speech-to-text, a small language model and a fast text-to-speech voice.
Which pieces did you pick, what hardware do they run on, and where did most of the latency come from?
Referenced repos
Indexed from Hugging Face and linked upstream, not hosted here.
ONNX exports of the Silero voice activity detector in fp32, fp16 and 8-bit forms (the int8 file is about 0.6 MB) that mark which short audio frames contain speech.
An 8M-parameter multilingual turn-detection model, a Whisper Tiny encoder with a linear head, that predicts from raw audio whether the speaker has finished; the int8 CPU ONNX build is about 8.7 MB.
The small size of the Moonshine Streaming family, an English-only encoder-decoder recognizer that uses a sliding-window encoder so it can transcribe live audio with low latency on edge hardware.
Smallest dense model in the Qwen3 family, with optional thinking mode and function-calling support.
An 82M-parameter text-to-speech model built on StyleTTS 2 components, shipped with a set of preset voices across several languages.