How much VRAM does ace-step-v1.5-1d-vae-stable-audio-format need?

BF16 is exactly 337,420,928 bytes (0.31 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

ACE-Step · text

ace-step-v1.5-1d-vae-stable-audio-format

ACE-Step/ace-step-v1.5-1d-vae-stable-audio-format

ace-step-v1.5-1d-vae-stable-audio-format at BF16 is exactly 337,420,928 bytes (0.31 GiB / 0.34 GB) — an effective 16.001 bits per weight, not the nominal 1.

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

Parameters

169M

Architecture

acestep-vae

null layers

Context

—

native (config.json)

License

mit

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
BF16	0.31 GiB	337,420,928	16.001	—	scragnog

Compare with

same modality, comparable size

vs Llama-3.2-1B-Instruct vs ced-base vs gemma-3-1b-it vs Qwen3-1.7B vs embeddinggemma-300m vs Qwen3-0.6B All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts BF16 at roughly 0.09 GiB. The real file is 0.31 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

—

Vocab

—

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does ace-step-v1.5-1d-vae-stable-audio-format need?: BF16 is exactly 337,420,928 bytes (0.31 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of ace-step-v1.5-1d-vae-stable-audio-format 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.