How much VRAM does SmolVLM-500M-Instruct need?

Q8_0 is exactly 436,806,912 bytes (0.41 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is SmolVLM-500M-Instruct's KV cache?

1.25 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.

HuggingFaceTB · text

SmolVLM-500M-Instruct

HuggingFaceTB/SmolVLM-500M-Instruct

SmolVLM-500M-Instruct at Q8_0 is exactly 436,806,912 bytes (0.41 GiB / 0.44 GB) — an effective 6.886 bits per weight, not the nominal 8. Its KV cache at 32K is 1.25 GiB.

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

Parameters

507M

Architecture

llama

32 layers

Context

8,192

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q8_0	0.41 GiB	436,806,912	6.886	291	ggml-org
F16	0.76 GiB	820,422,912	12.933	291	ggml-org

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.16 GiB	0.16 GiB	—	32 / 0 / 0
8,192	0.31 GiB	0.31 GiB	—	32 / 0 / 0
16,384	0.63 GiB	0.63 GiB	—	32 / 0 / 0
32,768	1.25 GiB	1.25 GiB	—	32 / 0 / 0
65,536	2.50 GiB	2.50 GiB	—	32 / 0 / 0
131,072	5.00 GiB	5.00 GiB	—	32 / 0 / 0

Compare with

same modality, comparable size

vs Qwen3-4B vs Llama-3.2-1B-Instruct vs ced-base vs gemma-3-1b-it vs Qwen3-1.7B vs embeddinggemma-300m All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q8_0 at roughly 0.27 GiB. The real file is 0.41 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

960

Vocab

49,280

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does SmolVLM-500M-Instruct need?: Q8_0 is exactly 436,806,912 bytes (0.41 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is SmolVLM-500M-Instruct's KV cache?: 1.25 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 SmolVLM-500M-Instruct 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.