RE

reazonspeech nemo v2

ReazonspeechGGUF by cstrUpdated 2026-07-12
Download on Hugging Face
Parameters619M
Context—
Layers—
Architectureparakeet
Smallest quant0.4 GB
Quick answer

reazonspeech nemo v2 needs about 1.2 GB of memory at Q8_0 (a 0.7 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 0.4 GB (Q4_K) to 1.2 GB (F16).

619 M‑parameter Japanese ASR model in GGUF format

reazonspeech‑nemo‑v2 is a Japanese automatic speech recognition model based on FastConformer‑RNN‑T. It supports long‑form audio up to several hours and uses a 3000‑token SentencePiece vocabulary. The GGUF files are available in Q4_K, Q8_0 and F16 quantisations.

Good at

  • Long‑form Japanese audio
  • FastConformer‑RNN‑T architecture
  • Local relative‑position attention

Limits

  • No vision support
  • No explicit context length limit provided
  • Apache‑2.0 licence

Best for

  • Transcribing long Japanese podcasts
  • Batch processing of Japanese audio files
  • Integrating with CrispASR CLI

Who it’s for. Developers building Japanese speech‑to‑text pipelines. Researchers needing a lightweight, long‑form ASR model for Japanese. Engineers integrating ASR into applications via the CrispASR library.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K0.4 GB 1.0 GB
Fits Get file →
Q8_00.7 GB 1.2 GB
Fits Get file →
F161.2 GB 1.9 GB
Fits Get file →

For 8 GB VRAM, the Q4_K file (0.41 GB) uses about 1 GB memory at 8 k context and is suitable for low‑budget inference. The Q8_0 file (0.67 GB) uses about 1.2 GB memory and offers higher precision. If memory is not a constraint, the F16 file (1.24 GB) uses about 1.9 GB memory and matches the official NeMo pipeline.

What GPU do you need for reazonspeech nemo v2?

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 F16 (1.2 GB file, about 1.9 GB in use).

Which reazonspeech nemo v2 quant should you choose?

Q8_0 (0.7 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.

If memory is tight, the smallest file is Q4_K at 0.4 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureparakeet
Parameters619M

Source

Quantized bycstr
Licenseapache-2.0
Downloads (HF)1,051
Last updated2026-07-12

Frequently asked questions about reazonspeech nemo v2

How much VRAM does reazonspeech nemo v2 need?

About 1.2 GB at Q8_0 with an 8K context window. That is the 0.7 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 reazonspeech nemo v2 on an 8 GB GPU?

Yes, with F16: the 1.2 GB file needs about 1.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of reazonspeech nemo v2 GGUF?

It depends on the quant: Q4_K is 0.4 GB, Q8_0 is 0.7 GB, F16 is 1.2 GB.

Where can I download reazonspeech nemo v2 GGUF?

The files are on Hugging Face in the cstr/reazonspeech-nemo-v2-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 reazonspeech nemo v2 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 this model on 8 GB VRAM?

Yes, the Q4_K quantisation requires about 1 GB memory at 8 k context, well within 8 GB VRAM.

Does this model support English?

No, the model is trained exclusively on Japanese audio and does not support other languages.

What is the recommended quantisation?

Q8_0 is recommended for general use; use F16 for the closest match to the official NeMo pipeline.

Other GGUFs of reazonspeech-nemo-v2

Similar models

Browse all →

More from cstr

All by cstr →

More Reazonspeech models

All Reazonspeech →

Other tiny models

All tiny models →