HO

Hojo ASR V1

HojoGGUF by cstrUpdated 2026-09-25
Download on Hugging Face
Parameters4.81B
Context—
Layers—
Architecturehojo_asr
Smallest quant4.4 GB
Quick answer

Hojo ASR V1 needs about 5.5 GB of memory at Q4_K (a 4.4 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 2 quants, from 4.4 GB (Q4_K) to 9.6 GB (F16).

Hojo ASR V1 – 4.81B parameter speech‑to‑text model

Hojo ASR V1 is a 4.81B parameter model designed for automatic speech recognition. It is available in GGUF format and supports multiple quantization levels. The model is provided by cstr and has no mixture‑of‑experts architecture.

Good at

  • Speech transcription
  • Low‑latency inference
  • Compact quantized models

Limits

  • No vision support
  • No license details provided
  • Context length not specified

Best for

  • Transcribing short audio clips
  • Real‑time voice assistants
  • Batch audio processing

Who it’s for. Developers building speech‑to‑text applications who need a lightweight, open‑source model. The model is suitable for edge devices with limited memory.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K4.4 GB 5.5 GB
Fits Get file →
F169.6 GB 11.3 GB
Too big Get file →

For 8 GB VRAM, use the Q4_K quantization (file size 4.42 GB, estimated memory 5.5 GB). For higher precision on 12 GB VRAM, use F16 (file size 9.64 GB, estimated memory 11.3 GB).

What GPU do you need for Hojo ASR V1?

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: no quant fits at an 8K context.
  • 6 GB VRAM: the best quant that fits is Q4_K (4.4 GB file, about 5.5 GB in use).
  • 8 GB VRAM: the best quant that fits is Q4_K (4.4 GB file, about 5.5 GB in use).
  • 12 GB VRAM: the best quant that fits is F16 (9.6 GB file, about 11.3 GB in use).

Which Hojo ASR V1 quant should you choose?

Q4_K (4.4 GB) is the middle option in this repository. 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

Architecturehojo_asr
Parameters4.81B

Source

Quantized bycstr
FamilyHojo
Downloads (HF)716
Last updated2026-09-25

Frequently asked questions about Hojo ASR V1

How much VRAM does Hojo ASR V1 need?

About 5.5 GB at Q4_K with an 8K context window. That is the 4.4 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 Hojo ASR V1 on an 8 GB GPU?

Yes, with Q4_K: the 4.4 GB file needs about 5.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Hojo ASR V1 GGUF?

It depends on the quant: Q4_K is 4.4 GB, F16 is 9.6 GB.

Where can I download Hojo ASR V1 GGUF?

The files are on Hugging Face in the cstr/Hojo-ASR-V1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run Hojo ASR V1 on 8 GB VRAM?

Yes, the Q4_K quantized model requires about 5.5 GB of VRAM at 8 k context, which fits within 8 GB.

What is the context length supported by Hojo ASR V1?

The context length is not specified in the provided information.

Does Hojo ASR V1 support vision tasks?

No, the model does not support vision capabilities.

Similar models

Browse all →

More from cstr

All by cstr →

More Hojo models

All Hojo →

Other small models

All small models →