How much VRAM does gemma-4-12B-it-qat-q4_0-unquantized-assistant need?

Q4_0 is exactly 322,913,920 bytes (0.30 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is gemma-4-12B-it-qat-q4_0-unquantized-assistant's KV cache?

0.29 GiB at 32K context with an f16 cache, computed per layer from the model's architecture — most of its layers use sliding-window attention, so a flat formula overstates this substantially. Quantizing the cache to q8_0 roughly halves it.

google · text

gemma-4-12B-it-qat-q4_0-unquantized-assistant

google/gemma-4-12B-it-qat-q4_0-unquantized-assistant

gemma-4-12B-it-qat-q4_0-unquantized-assistant at Q4_0 is exactly 322,913,920 bytes (0.30 GiB / 0.32 GB) — an effective 6.109 bits per weight, not the nominal 4. Its KV cache at 32K is 0.29 GiB, not the 1.00 GiB a flat formula predicts.

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

Parameters

423M

Architecture

gemma4-assistant

4 layers

Context

262,144

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_0	0.30 GiB	322,913,920	6.109	—	ysong21
Q8_0	0.43 GiB	465,126,880	8.800	—	RachidAR
Q8_0	0.43 GiB	465,127,040	8.800	—	Janvitos

KV cache by context

computed per layer — this model uses sliding-window attention

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.07 GiB	0.13 GiB	1.88×	1 / 3 / 0
8,192	0.10 GiB	0.25 GiB	2.56×	1 / 3 / 0
16,384	0.16 GiB	0.50 GiB	3.12×	1 / 3 / 0
32,768	0.29 GiB	1.00 GiB	3.51×	1 / 3 / 0
65,536	0.54 GiB	2.00 GiB	3.74×	1 / 3 / 0
131,072	1.04 GiB	4.00 GiB	3.86×	1 / 3 / 0

3 of 4 layers cache only a 1,024-token window rather than the full context, on a period of . Figures assume the default configuration; --swa-full disables the saving entirely.

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 Q4_0 at roughly 0.22 GiB. The real file is 0.30 GiB, because a quantization is a mixture and some tensors are always kept at higher precision. The larger discrepancy is the cache: a flat formula gives 1.00 GiB at 32K context where the real figure is 0.29 GiB, because most of this model's layers cache a fixed window rather than the whole context.

Architecture

from config.json

Layers

Attention heads

KV heads

Head dim

256

Hidden size

1024

Vocab

262,144

Sliding window

1024

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does gemma-4-12B-it-qat-q4_0-unquantized-assistant need?: Q4_0 is exactly 322,913,920 bytes (0.30 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is gemma-4-12B-it-qat-q4_0-unquantized-assistant's KV cache?: 0.29 GiB at 32K context with an f16 cache, computed per layer. Quantizing the cache to q8_0 roughly halves it, which is often the difference between a context length fitting and not.
Which quantization of gemma-4-12B-it-qat-q4_0-unquantized-assistant 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.