Moonshine Base needs about 0.6 GB of memory at F16 (a 0.1 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 0.1 GB (Q8_0) to 0.2 GB (F32).
Moonshine Base – 61.5M‑parameter English ASR model
Moonshine Base is a 61.5‑billion‑parameter encoder‑decoder transformer for English speech recognition. It accepts raw 16 kHz PCM audio and outputs transcript text. The model is released under the MIT license.
Good at
- Low‑memory ASR on constrained devices
- Fast inference with small context windows
- English‑only transcription
Limits
- No vision or translation support
- Requires 8 k context for memory estimates
- No timestamps in output
Best for
- Embedded speech‑to‑text on edge hardware
- Real‑time transcription in low‑resource apps
- Batch processing of short audio clips
Who it’s for. AI developers building English speech recognition systems for platforms with limited memory and compute. The model is suitable for developers who need a lightweight ASR solution without additional language or vision capabilities.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.1 GB | 0.6 GB | Fits | Get file → | |
| F16 | 0.1 GB | 0.6 GB | Fits | Get file → | |
| F32 | 0.2 GB | 0.8 GB | Fits | Get file → |
For 8 GB VRAM, use Q8_0 (0.08 GB file, ~0.6 GB memory at 8 k context). For 4 GB VRAM, F16 (0.13 GB file, ~0.6 GB memory) is a good balance. F32 (0.25 GB file, ~0.8 GB memory) requires more VRAM but offers the highest precision.
What GPU do you need for Moonshine Base?
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 F32 (0.2 GB file, about 0.8 GB in use).
Which Moonshine Base quant should you choose?
F16 (0.1 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
If memory is tight, the smallest file is Q8_0 at 0.1 GB, with the quality trade-offs that come with fewer bits. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Moonshine Base
How much VRAM does Moonshine Base need?
About 0.6 GB at F16 with an 8K context window. That is the 0.1 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 Moonshine Base on an 8 GB GPU?
Yes, with F32: the 0.2 GB file needs about 0.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Moonshine Base GGUF?
It depends on the quant: Q8_0 is 0.1 GB, F16 is 0.1 GB, F32 is 0.2 GB.
Where can I download Moonshine Base GGUF?
The files are on Hugging Face in the handy-computer/moonshine-base-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 Moonshine Base use?
Hugging Face lists the license as mit. 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 Moonshine Base on 4 GB VRAM?
Yes, the F16 quantization uses about 0.6 GB memory at 8 k context, fitting within 4 GB VRAM.
Does the model support timestamps?
No, Moonshine Base outputs transcript text only and does not provide timestamps.
Is the model multilingual?
No, it is English‑only and does not perform language detection or translation.