How much VRAM does gemma-4-26B-A4B-it-DFlash need?

Q4_K_M is exactly 266,221,536 bytes (0.25 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is gemma-4-26B-A4B-it-DFlash's KV cache?

0.16 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.

z-lab · text

gemma-4-26B-A4B-it-DFlash

z-lab/gemma-4-26B-A4B-it-DFlash

gemma-4-26B-A4B-it-DFlash at Q4_K_M is exactly 266,221,536 bytes (0.25 GiB / 0.27 GB) — an effective 4.957 bits per weight, not the nominal 4. Its KV cache at 32K is 0.16 GiB, not the 0.63 GiB a flat formula predicts.

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

Parameters

430M

Architecture

dflash

5 layers

Context

262,144

native (config.json)

License

—

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q2_K	0.16 GiB	174,196,704	3.243	—	Anbeeld
Q3_K_M	0.21 GiB	221,370,336	4.122	—	Anbeeld
IQ4_XS	0.23 GiB	246,431,808	4.588	—	williamliao
Q4_K_M	0.25 GiB	266,221,536	4.957	—	Alittlehammmer
Q4_K_M	0.25 GiB	267,270,208	4.976	—	williamliao
Q4_K_M	0.25 GiB	267,271,136	4.976	—	Anbeeld
Q5_K	0.29 GiB	315,242,464	5.869	—	Alittlehammmer
Q5_K_M	0.29 GiB	316,291,136	5.889	—	williamliao
Q5_K_M	0.29 GiB	316,292,064	5.889	—	Anbeeld
Q6_K	0.34 GiB	367,327,200	6.839	—	Alittlehammmer
Q6_K	0.34 GiB	368,375,872	6.859	—	williamliao
Q6_K	0.34 GiB	368,376,800	6.859	—	Anbeeld
Q8_0	0.44 GiB	471,384,032	8.776	—	Alittlehammmer
Q8_0	0.44 GiB	472,432,704	8.796	—	williamliao
Q8_0	0.44 GiB	472,433,632	8.796	—	Anbeeld
BF16	0.81 GiB	874,184,672	16.276	—	Alittlehammmer
F16	0.82 GiB	875,233,344	16.295	—	williamliao
BF16	0.82 GiB	875,234,272	16.295	—	Anbeeld

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	0.05 GiB	0.08 GiB	1.43×	1 / 4 / 0
8,192	0.07 GiB	0.16 GiB	2.22×	1 / 4 / 0
16,384	0.10 GiB	0.31 GiB	3.08×	1 / 4 / 0
32,768	0.16 GiB	0.63 GiB	3.81×	1 / 4 / 0
65,536	0.29 GiB	1.25 GiB	4.32×	1 / 4 / 0
131,072	0.54 GiB	2.50 GiB	4.64×	1 / 4 / 0

4 of 5 layers cache only a 2,048-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 embeddinggemma-300m vs Qwen3-0.6B vs Qwen2.5-0.5B-Instruct vs SmolLM2-135M-Instruct All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 0.23 GiB. The real file is 0.25 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 0.63 GiB at 32K context where the real figure is 0.16 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

128

Hidden size

2816

Vocab

262,144

Sliding window

2048

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

true

Questions people ask

How much VRAM does gemma-4-26B-A4B-it-DFlash need?: Q4_K_M is exactly 266,221,536 bytes (0.25 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-4-26B-A4B-it-DFlash's KV cache?: 0.16 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-4-26B-A4B-it-DFlash 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.