Kokoro-ONNX (ACut multi-language hub)

Single Hugging Face repo for ACut desktop Kokoro TTS assets.

  • Voicepacks (.npy): downloaded on demand by the app (keeps the installer small).
  • Vietnamese ONNX models: under vi/ for packaging / rebuilds.
  • 82M ONNX: prepared from thewh1teagle/kokoro-onnx releases (kokoro-v1.0.int8.onnx); voicepacks live here under 82m/.

Layout

Path Contents
vi/kokoro_vi.int8.onnx Shipped INT8 model (~79 MB)
vi/kokoro_vi.fp16.onnx FP16 weight-pack (~156 MB)
vi/kokoro_vi.fp32.onnx Upstream FP32 (~311 MB)
vi/config.json Model config
vi/voicepacks/*.npy Runtime style packs (14)
vi/voicepacks/*.pt Upstream .pt style packs
vi/voices-manifest.json VI catalog
82m/voicepacks/*.npy Kokoro-82M voices (54)
82m/voices-manifest.json 82M catalog

Download examples

# Vietnamese voicepacks only
hf download vongocanhthi/Kokoro-ONNX vi/voicepacks --local-dir ./kokoro-vi-voicepacks

# 82M voicepacks only
hf download vongocanhthi/Kokoro-ONNX 82m/voicepacks --local-dir ./kokoro-82m-voicepacks

Sources

  • Vietnamese: contextboxai/Kokoro-Vietnamese (Apache-2.0)
  • 82M voices: hexgrad Kokoro-82M via thewh1teagle/kokoro-onnx voices-v1.0.bin (split to per-voice .npy)

Try ACut

ACut — privacy-first video editor (browser + desktop): https://www.acutapp.com/

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support