ME

MedASR

MedasrGGUF by handy-computerUpdated 2026-10-16
Download on Hugging Face
Parameters105M
Context—
Layers—
Architecturemedasr
Smallest quant0.1 GB
Quick answer

MedASR needs about 0.6 GB of memory at Q4_K_M (a 0.1 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 6 quants, from 0.1 GB (Q4_K_M) to 0.4 GB (F32).

MedASR: Medical speech‑to‑text model in GGUF format

MedASR is a 0.105‑billion‑parameter Conformer encoder optimized for medical dictation. It uses a 512‑token SentencePiece CTC head and greedy CTC decoding. The model is available in several GGUF quantizations.

Good at

  • Medical dictation
  • Low‑latency transcription
  • CTC‑based decoding

Limits

  • No vision support
  • Requires 16 kHz mono WAV
  • License: Health AI Developer Foundations

Best for

  • Radiology dictation
  • Internal medicine notes
  • Family medicine transcripts

Who it’s for. Developers building offline medical transcription tools. Researchers needing a lightweight, domain‑specific ASR model for clinical data. Teams that require deterministic greedy decoding without beam search.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M0.1 GB 0.6 GB
Fits Get file →
Q5_K_M0.1 GB 0.6 GB
Fits Get file →
Q6_K0.1 GB 0.6 GB
Fits Get file →
Q8_00.1 GB 0.6 GB
Fits Get file →
F160.2 GB 0.7 GB
Fits Get file →
F320.4 GB 1.0 GB
Fits Get file →

For 8 GB VRAM, use Q8_0 (0.13 GB, 0.6 GB memory at 8k context). If memory is tighter, Q5_K_M (0.09 GB, 0.6 GB memory) is an acceptable trade‑off. Avoid Q4_K_M due to measurable WER degradation.

What GPU do you need for MedASR?

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 (0.4 GB file, about 1.0 GB in use).

Which MedASR quant should you choose?

Q4_K_M is the usual starting point for MedASR: it balances file size and quality well. Here it is a 0.1 GB file.

The Q8_0 file is 0.1 GB, about 1.5 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.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturemedasr
Parameters105M

Source

Quantized byhandy-computer
FamilyMedasr
Base modelgoogle/medasr
Licenseother
Downloads (HF)1,401
Last updated2026-10-16

Frequently asked questions about MedASR

How much VRAM does MedASR need?

About 0.6 GB at Q4_K_M 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 MedASR on an 8 GB GPU?

Yes, with F32: the 0.4 GB file needs about 1.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of MedASR GGUF?

It depends on the quant: Q4_K_M is 0.1 GB, Q5_K_M is 0.1 GB, Q6_K is 0.1 GB, Q8_0 is 0.1 GB, F16 is 0.2 GB, F32 is 0.4 GB.

Where can I download MedASR GGUF?

The files are on Hugging Face in the handy-computer/medasr-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 MedASR 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 MedASR on 8 GB VRAM?

Yes, the Q8_0 quantization requires about 0.6 GB of VRAM at 8k context, well within 8 GB.

Does MedASR support multi‑language transcription?

No, the model is trained for English medical dictation only.

What is the recommended quantization for best accuracy?

Q8_0 is the recommended default; Q5_K_M is a smaller alternative with minimal accuracy loss.

Other GGUFs of medasr

Similar models

Browse all →

More from handy-computer

All by handy-computer →

More Medasr models

All Medasr →

Other tiny models

All tiny models →