How much VRAM does PaddleOCR-VL-1.5 need?

Q4_K_M is exactly 300,218,208 bytes (0.28 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is PaddleOCR-VL-1.5's KV cache?

0.56 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.

PaddlePaddle · text

PaddleOCR-VL-1.5

PaddlePaddle/PaddleOCR-VL-1.5

PaddleOCR-VL-1.5 at Q4_K_M is exactly 300,218,208 bytes (0.28 GiB / 0.30 GB) — an effective 2.506 bits per weight, not the nominal 4. Its KV cache at 32K is 0.56 GiB.

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

Parameters

959M

Architecture

paddleocr

18 layers

Context

131,072

native (config.json)

License

—

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
IQ4_NL	0.25 GiB	265,598,048	2.217	—	Mungert
Q3_K_S	0.25 GiB	271,950,944	2.270	—	Mungert
IQ3_XXS	0.26 GiB	274,244,704	2.289	—	Mungert
IQ3_XS	0.26 GiB	280,437,856	2.340	—	Mungert
IQ3_M	0.27 GiB	293,921,888	2.453	—	Mungert
Q4_K_M	0.28 GiB	300,218,208	2.506	—	octopusmegalopod
Q3_K_M	0.28 GiB	303,981,664	2.537	—	Mungert
Q4_1	0.29 GiB	307,410,016	2.566	—	Mungert
IQ4_XS	0.29 GiB	312,390,752	2.607	—	Mungert
Q4_K_S	0.30 GiB	326,431,840	2.724	—	Mungert
Q4_0	0.35 GiB	370,914,400	3.095	—	Mungert
Q4_K_M	0.36 GiB	382,084,192	3.189	—	Mungert
Q5_0	0.38 GiB	402,764,896	3.361	—	Mungert
Q5_K_M	0.38 GiB	411,747,424	3.436	—	Mungert
Q5_1	0.39 GiB	418,690,144	3.494	—	Mungert
Q6_K_M	0.41 GiB	436,606,048	3.644	—	Mungert
Q8_0	0.46 GiB	498,315,104	4.159	—	octopusmegalopod
Q8_0	0.46 GiB	498,316,064	4.159	—	Mungert
Q8_0	0.46 GiB	498,316,224	4.159	—	noctrex
BF16	0.87 GiB	935,768,864	7.810	—	Mungert
BF16	0.87 GiB	935,769,024	7.810	—	noctrex
F16	0.87 GiB	935,769,024	7.810	—	noctrex

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.07 GiB	0.07 GiB	—	18 / 0 / 0
8,192	0.14 GiB	0.14 GiB	—	18 / 0 / 0
16,384	0.28 GiB	0.28 GiB	—	18 / 0 / 0
32,768	0.56 GiB	0.56 GiB	—	18 / 0 / 0
65,536	1.13 GiB	1.13 GiB	—	18 / 0 / 0
131,072	2.25 GiB	2.25 GiB	—	18 / 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 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_K_M at roughly 0.50 GiB. The real file is 0.28 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

103,424

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does PaddleOCR-VL-1.5 need?: Q4_K_M is exactly 300,218,208 bytes (0.28 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is PaddleOCR-VL-1.5's KV cache?: 0.56 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 PaddleOCR-VL-1.5 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.