How much VRAM does moonshine-base need?

Q4_K is exactly 46,923,872 bytes (0.04 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

UsefulSensors · audio asr

moonshine-base

UsefulSensors/moonshine-base

moonshine-base at Q4_K is exactly 46,923,872 bytes (0.04 GiB / 0.05 GB) — an effective 6.103 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)From the file· KV per layer

Parameters

62M

Architecture

moonshine

null layers

Context

194

native (config.json)

License

mit

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_K	0.04 GiB	46,923,872	6.103	—	cstr
Q8_0	0.07 GiB	75,890,784	9.870	—	cstr
Q8_0	0.07 GiB	77,476,480	10.076	—	handy-computer
F16	0.12 GiB	131,789,440	17.140	—	handy-computer
F32	0.23 GiB	247,657,088	32.208	—	handy-computer

Compare with

same modality, comparable size

vs whisper-small vs whisper-base vs whisper-tiny vs parakeet-tdt_ctc-110m vs whisper-base.en vs moonshine-streaming-tiny All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K at roughly 0.03 GiB. The real file is 0.04 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json

Layers

—

Attention heads

—

KV heads

—

Head dim

—

Hidden size

416

Vocab

32,768

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does moonshine-base need?: Q4_K is exactly 46,923,872 bytes (0.04 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of moonshine-base should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.