How much VRAM does LightOnOCR-2-1B-bbox-soup need?

Q4_K_M is exactly 396,701,728 bytes (0.37 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is LightOnOCR-2-1B-bbox-soup's KV cache?

3.50 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.

lightonai · text

LightOnOCR-2-1B-bbox-soup

lightonai/LightOnOCR-2-1B-bbox-soup

LightOnOCR-2-1B-bbox-soup at Q4_K_M is exactly 396,701,728 bytes (0.37 GiB / 0.40 GB) — an effective 3.156 bits per weight, not the nominal 4. Its KV cache at 32K is 3.50 GiB.

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

Parameters

1.0B

Architecture

qwen3

28 layers

Context

16,384

native (config.json)

License

—

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
IQ4_XS	0.34 GiB	367,800,352	2.926	—	noctrex
IQ4_NL	0.36 GiB	381,562,912	3.035	—	noctrex
Q4_K_M	0.37 GiB	396,701,728	3.156	—	noctrex
Q5_K_M	0.41 GiB	444,411,936	3.535	—	noctrex
Q6_K	0.46 GiB	495,104,032	3.939	—	noctrex
Q8_0	0.60 GiB	639,443,776	5.087	—	noctrex
F16	1.12 GiB	1,198,179,136	9.532	—	noctrex
BF16	1.12 GiB	1,198,179,136	9.532	—	noctrex

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.44 GiB	0.44 GiB	—	28 / 0 / 0
8,192	0.88 GiB	0.88 GiB	—	28 / 0 / 0
16,384	1.75 GiB	1.75 GiB	—	28 / 0 / 0
32,768	3.50 GiB	3.50 GiB	—	28 / 0 / 0
65,536	7.00 GiB	7.00 GiB	—	28 / 0 / 0
131,072	14.00 GiB	14.00 GiB	—	28 / 0 / 0

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 Llama-3.2-3B-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 0.53 GiB. The real file is 0.37 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

128

Hidden size

1024

Vocab

151,936

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

false

Questions people ask

How much VRAM does LightOnOCR-2-1B-bbox-soup need?: Q4_K_M is exactly 396,701,728 bytes (0.37 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is LightOnOCR-2-1B-bbox-soup's KV cache?: 3.50 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 LightOnOCR-2-1B-bbox-soup 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.