sayit asr needs about 1.0 GB of memory at Q4_K_M (a 0.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 7 quants, from 0.3 GB (Q8_0) to 1.5 GB (Q5_K_M).
SayIt ASR: lightweight local speech recognition in GGUF format
SayIt ASR provides local speech recognition via GGUF files converted for transcribe.cpp. It is based on Fun-ASR-Nano-2512 and supports 31 languages with real‑time transcription. The model runs without GPU or Python, using a single binary.
Good at
- Real‑time transcription
- Low‑latency edge deployment
- Multi‑language support
Limits
- Requires transcribe.cpp runtime
- No GPU acceleration
- No vision or VAD features in GGUF
Best for
- Desktop voice typing
- Embedded transcription devices
- Offline meeting note capture
Who it’s for. Users needing quick, local speech recognition on consumer hardware. Ideal for developers integrating voice input into Windows applications without external dependencies.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.3 GB | 0.8 GB | Fits | Get file → | |
| Q4_K_M | 0.5 GB | 1.0 GB | Fits | Get file → | |
| Q4_K_M | 0.5 GB | 1.1 GB | Fits | Get file → | |
| Q8_0 | 0.9 GB | 1.5 GB | Fits | Get file → | |
| Q8_0 | 0.9 GB | 1.5 GB | Fits | Get file → | |
| Q4_K_M | 1.3 GB | 2.0 GB | Fits | Get file → | |
| Q5_K_M | 1.5 GB | 2.2 GB | Fits | Get file → |
For 8 GB VRAM, use Q8_0 (0.25 GB) or Q4_K_M (0.48 GB) files; memory usage is approximately 0.8–1 GB at 8 k context. For 4 GB VRAM, choose Q4_K_M (0.48 GB) or Q4_K_M (0.5 GB) which use about 1 GB memory. Larger models (1.32 GB, 1.52 GB) require 2–2.2 GB memory and are suitable for systems with 8 GB or more.
What GPU do you need for sayit asr?
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 Q5_K_M (1.5 GB file, about 2.2 GB in use).
Which sayit asr quant should you choose?
Q4_K_M is the usual starting point for sayit asr: it balances file size and quality well. Here it is a 0.5 GB file.
The Q8_0 file is 0.3 GB, about 0.5 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q8_0 at 0.3 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 sayit asr
How much VRAM does sayit asr need?
About 1.0 GB at Q4_K_M with an 8K context window. That is the 0.5 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 sayit asr on an 8 GB GPU?
Yes, with Q5_K_M: the 1.5 GB file needs about 2.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of sayit asr GGUF?
It depends on the quant: Q8_0 is 0.3 GB, Q4_K_M is 0.5 GB, Q4_K_M is 0.5 GB, Q8_0 is 0.9 GB, Q8_0 is 0.9 GB, Q4_K_M is 1.3 GB and 1 more.
Where can I download sayit asr GGUF?
The files are on Hugging Face in the cswk/sayit-asr-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 sayit asr use?
Hugging Face lists the license as other. 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 SayIt ASR on 4 GB VRAM?
Yes, use the Q4_K_M 0.48 GB or 0.5 GB file; estimated memory is about 1 GB at 8 k context.
Does the model support GPU acceleration?
No, the GGUF files are designed for transcribe.cpp and run on CPU only.
What runtime is required to load these GGUF files?
They require transcribe.cpp, not llama.cpp or whisper.cpp; the runtime must match the architecture tags.