How much VRAM does MinerU2.5-Pro-2605-1.2B need?

Q4_K_M is exactly 397,806,688 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 MinerU2.5-Pro-2605-1.2B's KV cache?

0.38 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.

opendatalab · text

MinerU2.5-Pro-2605-1.2B

opendatalab/MinerU2.5-Pro-2605-1.2B

MinerU2.5-Pro-2605-1.2B at Q4_K_M is exactly 397,806,688 bytes (0.37 GiB / 0.40 GB) — an effective 2.753 bits per weight, not the nominal 4. Its KV cache at 32K is 0.38 GiB.

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

Parameters

1.2B

Architecture

qwen2vl

24 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
Q3_K_S	0.32 GiB	338,262,112	2.341	—	mradermacher
Q2_K	0.32 GiB	338,606,176	2.343	—	mradermacher
IQ4_XS	0.33 GiB	351,444,064	2.432	—	mradermacher
Q3_K_M	0.33 GiB	355,465,312	2.460	—	mradermacher
Q3_K_L	0.34 GiB	369,356,896	2.556	—	mradermacher
Q4_K_S	0.36 GiB	385,470,560	2.668	—	mradermacher
Q4_K_M	0.37 GiB	397,806,688	2.753	—	mradermacher
Q5_K_S	0.38 GiB	412,708,960	2.856	—	mradermacher
Q5_K_M	0.39 GiB	420,084,832	2.907	—	mradermacher
Q6_K	0.47 GiB	505,735,264	3.500	—	mradermacher
Q8_0	0.49 GiB	531,066,496	3.675	—	JamePeng2023
Q8_0	0.49 GiB	531,066,976	3.675	—	mradermacher
BF16	0.93 GiB	994,155,136	6.880	—	JamePeng2023
F16	0.93 GiB	994,155,616	6.880	—	mradermacher

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.05 GiB	0.05 GiB	—	24 / 0 / 0
8,192	0.09 GiB	0.09 GiB	—	24 / 0 / 0
16,384	0.19 GiB	0.19 GiB	—	24 / 0 / 0
32,768	0.38 GiB	0.38 GiB	—	24 / 0 / 0
65,536	0.75 GiB	0.75 GiB	—	24 / 0 / 0
131,072	1.50 GiB	1.50 GiB	—	24 / 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 0.61 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

Hidden size

896

Vocab

151,936

Sliding window

32768

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

false

This model declares a sliding window but sets use_sliding_window: false, so the window is not applied. Honouring the field without the flag understates KV for the whole family.

Questions people ask

How much VRAM does MinerU2.5-Pro-2605-1.2B need?: Q4_K_M is exactly 397,806,688 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 MinerU2.5-Pro-2605-1.2B's KV cache?: 0.38 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 MinerU2.5-Pro-2605-1.2B 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.