runtime-llamacpp-v0.1.2

diffusionstudio/coreruntime-llamacpp-v0.1.2Jun 21, 2026by LauraGPT

AI Summary

Prebuilt binaries for running SenseVoice ASR locally using the llama.cpp runtime, supporting Chinese and Cantonese without requiring Python or a build process.

Key Highlights

Support for SenseVoice (and Paraformer / Fun-ASR-Nano) with built-in FSMN-VAD
New q8 GGUF models that are approximately half the size of f16 with similar accuracy
No Python dependency or build required
Cross-platform support (Linux x64/arm64, macOS arm64, Windows x64)

New Features

SenseVoice ASR integration
q8 GGUF model support
On-device whisper.cpp-style ASR
Cross-platform binary distribution

Full Release Notes

Prebuilt self-contained binaries for running **SenseVoice** (and Paraformer / Fun-ASR-Nano) locally with the FunASR llama.cpp / GGUF runtime — built-in FSMN-VAD, whisper.cpp-style on-device ASR, strong on Chinese & Cantonese.

**New:** q8 GGUF models are ~half the size of f16 with the same accuracy.

```bash
bash download-funasr-model.sh sensevoice ./gguf
llama-funasr-sensevoice -m ./gguf/sensevoice-small-q8.gguf --vad ./gguf/fsmn-vad.gguf -a audio.wav
```

No Python, no build. Linux (x64/arm64), macOS (arm64), Windows (x64). Docs: `runtime/llama.cpp/README.md`.