MOSS Audio 4B Instruct needs about 5.1 GB of memory at Q4_K (a 4.1 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 3 quants, from 0.0 GB to 10.5 GB (F16).
MOSS Audio 4B Instruct – open‑source audio‑understanding model
MOSS Audio 4B Instruct is a 5.22B parameter model designed for unified audio understanding. It supports speech recognition, environmental sound analysis, music understanding, audio captioning, time‑aware QA, and complex reasoning. The model is released under an Apache‑2.0 license.
Good at
- Speech recognition with timestamps
- Environmental sound and music analysis
- Time‑aware question answering
Limits
- No vision support
- Requires 8k context for full functionality
- Large memory for higher precision quantizations
Best for
- Transcribing podcasts with speaker tags
- Generating audio summaries for meetings
- Answering time‑specific audio queries
Who it’s for. Developers building audio‑centric applications who need a single model for speech, sound, and music understanding. Researchers exploring multimodal audio reasoning can also benefit.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 0.5 GB | Fits | Get file → | |
| Q4_K | 4.1 GB | 5.1 GB | Fits | Get file → | |
| F16 | 10.5 GB | 12.2 GB | Too big | Get file → |
For 8 GB VRAM use the Q4_K quantization (4.11 GB file, ~5.1 GB memory at 8k context). If more VRAM is available, the F16 quantization (10.45 GB file, ~12.2 GB memory) offers higher precision. The unknown quant is very small (0.02 GB) but memory usage is not specified.
What GPU do you need for MOSS Audio 4B Instruct?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 6 GB VRAM: the best quant that fits is Q4_K (4.1 GB file, about 5.1 GB in use).
- 8 GB VRAM: the best quant that fits is Q4_K (4.1 GB file, about 5.1 GB in use).
- 12 GB VRAM: the best quant that fits is Q4_K (4.1 GB file, about 5.1 GB in use).
- 16 GB VRAM: the best quant that fits is F16 (10.5 GB file, about 12.2 GB in use).
Which MOSS Audio 4B Instruct quant should you choose?
Q4_K (4.1 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about MOSS Audio 4B Instruct
How much VRAM does MOSS Audio 4B Instruct need?
About 5.1 GB at Q4_K with an 8K context window. That is the 4.1 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run MOSS Audio 4B Instruct on an 8 GB GPU?
Yes, with Q4_K: the 4.1 GB file needs about 5.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of MOSS Audio 4B Instruct GGUF?
It depends on the quant: file is 0.0 GB, Q4_K is 4.1 GB, F16 is 10.5 GB.
Where can I download MOSS Audio 4B Instruct GGUF?
The files are on Hugging Face in the cstr/MOSS-Audio-4B-Instruct-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does MOSS Audio 4B Instruct use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run MOSS Audio 4B Instruct on 8 GB VRAM?
Yes, the Q4_K quantization requires about 5.1 GB of VRAM at 8k context, which fits within 8 GB.
Does the model support Mandarin speech?
Yes, it supports Mandarin and English speech recognition.
Is there a vision component in this model?
No, the model is audio‑only and does not include vision capabilities.