ModelListedExternal

Fun ASR Nano GGUF

GGUF build of Fun-ASR-Nano (SenseVoice SAN-M encoder + adaptor + Qwen3-0.6B LLM decoder) for the zero-Python, CPU/edge FunASR llama.cpp runtime — the accuracy leader (LLM decoder), single C++ binary.

Sourcehuggingface.co/FunAudioLLM/Fun-ASR-Nano-GGUF/tree/46e849502a867080d66d351b8dfb1018b607e509

KindOtherVersionRevision 46e8495Bgrade

Model source

Kind: Other
File size: 448–767 MB
Version: vRevision 46e8495
Updated: Oct 26, 2026
Downloads: 10.2K
Source: Hugging Face

Source description

Fun-ASR-Nano · GGUF (FunASR llama.cpp runtime)

GGUF build of Fun-ASR-Nano (SenseVoice SAN-M encoder + adaptor + Qwen3-0.6B LLM decoder) for the zero-Python, CPU/edge FunASR llama.cpp runtime — the accuracy leader (LLM decoder), single C++ binary.

LLM quantization (pick by size vs accuracy)

The Fun-ASR-Nano LLM (Qwen3-0.6B) ships in three tiers — all within 0.1% CER (184-file micro-CER). Pair any with funasr-encoder-f16.gguf (470 MB).

LLM file	size	CER ↓	speed
`qwen3-0.6b-q4km.gguf`	484 MB	8.35%	6.1×
`qwen3-0.6b-q5km.gguf`	551 MB	8.25%	5.7×
`qwen3-0.6b-q8_0.gguf`	805 MB	8.30%	6.0×

Recommended: q4_K_M (smallest) or q5_K_M (best).

Get it running (no Python, no build)

These are GGUF weights for the FunASR llama.cpp runtime — a whisper.cpp-style, single self-contained binary for CPU / edge. Grab a prebuilt binary, then fetch this model and run:

Prebuilt binaries (Linux / macOS / Windows) → GitHub Releases (tag runtime-llamacpp-v*)
Deployment guide & qualified benchmarks → funasr.com/deploy/llama-cpp

bash download-funasr-model.sh nano ./gguf
llama-funasr-cli --enc ./gguf/funasr-encoder-f16.gguf -m ./gguf/qwen3-0.6b-q8_0.gguf --vad ./gguf/fsmn-vad.gguf -a audio.wav

Files

file	size	notes
`funasr-encoder-f16.gguf`	470 MB	audio encoder + adaptor (f16)
`qwen3-0.6b-q8_0.gguf`	805 MB	LLM decoder, recommended (Q8_0)
`qwen3-0.6b-q4km.gguf`	484 MB	LLM decoder, smaller (Q4_K_M)

Usage (needs both the encoder and the LLM gguf)

llama-funasr-cli --enc funasr-encoder-f16.gguf -m qwen3-0.6b-q8_0.gguf -a audio.wav --vad fsmn-vad.gguf

On CPU: 8.30 % CER on the 184-clip Mandarin benchmark (vs whisper.cpp 22–31 %).

Links

🧩 Runtime & build: Fun-ASR · runtime/llama.cpp — ⭐ Star Fun-ASR!
Source model: FunAudioLLM/Fun-ASR-Nano-2512

Sources

1 source

Hugging FacePrimaryVersion 46e849502a867080d66d351b8dfb1018b607e509

huggingface.co/FunAudioLLM/Fun-ASR-Nano-GGUF/tree/46e849502a867080d66d351b8dfb1018b607e509

Verified Oct 26

Model artifacts

4 artifacts

funasr-encoder-f16.gguf

gguf · 448 MB · SHA-256 f92f91d01a24…e8a2 · Hugging Face

Download

qwen3-0.6b-q4km.gguf

gguf · 462 MB · SHA-256 cc5057552aa9…6f22 · Hugging Face

Download

qwen3-0.6b-q5km.gguf

gguf · 526 MB · SHA-256 dc2e6e195c53…a2a2 · Hugging Face

Download

qwen3-0.6b-q8_0.gguf

gguf · 767 MB · SHA-256 819f385dc0e0…bf27 · Hugging Face

Download

Source excerpts

3 excerpts

ReadmeAug 7huggingface.co/FunAudioLLM/Fun-ASR-Nano-GGUF

--- license: apache-2.0 language: - zh - en library_name: gguf tags: - automatic-speech-recognition - asr - fun-asr - funasr - qwen3 - llama.cpp - ggml - cpu - chinese - arxiv:2407.04051 pipeline_tag: automatic-speech-recognition --- # Fun-ASR-Nano · GGUF (FunASR llama.cpp runtime) GGUF build of **Fun-ASR-Nano** (SenseVoice SAN-M encoder + adaptor + **Qwen3-0.6B** LLM decoder) for the zero-Python, CPU/edge **[FunASR llama.cpp runtime](https://github.com/FunAudioLLM/Fun-ASR/tree/main/runtime/llama.cpp)** — the accuracy leader (LLM decoder), single C++ binary. ## LLM quantization (pick by size vs accuracy) The Fun-ASR-Nano LLM (Qwen3-0.6B) ships in three tiers — all within 0.1% CER (184-file micro-CER). Pair any with `funasr-encoder-f16.gguf` (470 MB). | LLM file | size | CER ↓ | speed | |---|---|---|---| | `qwen3-0.6b-q4km.gguf` | **484 MB** | 8.35% | 6.1× | smallest | | `qwen3-0.6b-q5km.gguf` | 551 MB | **8.25%** | 5.7× | best accuracy | | `qwen3-0.6b-q8_0.gguf` | 805 MB | 8.30% | 6.0× | | Recommended: **q4_K_M** (smallest) or **q5_K_M** (best). ## Get it running (no Python, no build) These are GGUF weights for the **[FunASR llama.cpp runtime](https://github.com/modelscope/FunASR/tree/main/runtime/llama.cpp)** — a whisper.cpp-style, single self-contained binary for CPU / edge. Grab a prebuilt binary, then fetch this model and run: - **Prebuilt binaries (Linux / macOS / Windows) → [GitHub Releases](https://github.com/modelscope/FunASR/releases)** (tag `runtime-llamacpp-v*`) - **Deployment guide & qualified benchmarks → [funasr.com/deploy/llama-cpp](https://www.funasr.com/deploy/llama-cpp.html)** ```bash bash download-funasr-model.sh nano ./gguf llama-funasr-cli --enc ./gguf/funasr-encoder-f16.gguf -m ./gguf/qwen3-0.6b-q8_0.gguf --vad ./gguf/fsmn-vad.gguf -a audio.wav ``` ## Files | file | size | notes | |---|---|---| | `funasr-encoder-f16.gguf` | 470 MB | audio encoder + adaptor (f16) | | `qwen3-0.6b-q8_0.gguf` | 805 MB | LLM decoder, **recommended** (Q8_0) | | `qwen3-0.6b-q4km.gguf` | 484 MB | LLM decoder, smaller (Q4_K_M) | ## Usage (needs both the encoder and the LLM gguf) ```bash llama-funasr-cli --enc funasr-encoder-f16.gguf -m qwen3-0.6b-q8_0.gguf -a audio.wav --vad fsmn-vad.gguf ``` On CPU: **8.30 % CER** on the 184-clip Mandarin benchmark (vs whisper.cpp 22–31 %). ## Links - 🧩 Runtime & build: **[Fun-ASR · runtime/llama.cpp](https://g...

NoteOct 26huggingface.co/FunAudioLLM/Fun-ASR-Nano-GGUF

Source context: 10196 downloads · 6 likes · Pipeline automatic-speech-recognition · Library gguf · Repo FunAudioLLM/Fun-ASR-Nano-GGUF

NoteAug 7huggingface.co/FunAudioLLM/Fun-ASR-Nano-GGUF

Source context: 3717 downloads · 3 likes · Pipeline automatic-speech-recognition · Library gguf · Repo FunAudioLLM/Fun-ASR-Nano-GGUF