How much VRAM does granite-vision-3.3-2b need?

Q4_K_M is exactly 1,545,294,400 bytes (1.44 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is granite-vision-3.3-2b's KV cache?

2.50 GiB at 32K context with an f16 cache, computed per layer from the model's architecture. Quantizing the cache to q8_0 roughly halves it.

ibm-granite · text

granite-vision-3.3-2b

ibm-granite/granite-vision-3.3-2b

granite-vision-3.3-2b at Q4_K_M is exactly 1,545,294,400 bytes (1.44 GiB / 1.55 GB) — an effective 4.155 bits per weight, not the nominal 4. Its KV cache at 32K is 2.50 GiB.

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

Parameters

3.0B

Architecture

granite

40 layers

Context

131,072

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_K_M	1.44 GiB	1,545,294,400	4.155	—	ibm-granite
Q5_K_M	1.68 GiB	1,804,816,960	4.853	—	ibm-granite
Q4_K	1.87 GiB	2,007,964,352	5.399	—	cstr
Q6_K	1.94 GiB	2,080,559,680	5.594	—	ibm-granite
Q8_0	2.51 GiB	2,694,109,696	7.244	—	ibm-granite
Q8_0	3.14 GiB	3,369,727,680	9.060	—	cstr
BF16	4.72 GiB	5,069,142,016	13.630	—	ibm-granite
F16	5.52 GiB	5,923,033,824	15.925	—	cstr

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.31 GiB	0.31 GiB	—	40 / 0 / 0
8,192	0.63 GiB	0.63 GiB	—	40 / 0 / 0
16,384	1.25 GiB	1.25 GiB	—	40 / 0 / 0
32,768	2.50 GiB	2.50 GiB	—	40 / 0 / 0
65,536	5.00 GiB	5.00 GiB	—	40 / 0 / 0
131,072	10.00 GiB	10.00 GiB	—	40 / 0 / 0

Compare with

same modality, comparable size

vs gemma-4-E4B-it vs Qwen3-4B vs Qwen3-8B vs HyperCLOVAX-SEED-Text-Instruct-1.5B vs Llama-3.2-1B-Instruct vs Qwen3-VL-8B-Instruct-abliterated-v1 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 1.56 GiB. The real file is 1.44 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

2048

Vocab

49,156

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does granite-vision-3.3-2b need?: Q4_K_M is exactly 1,545,294,400 bytes (1.44 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is granite-vision-3.3-2b's KV cache?: 2.50 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 granite-vision-3.3-2b 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.