How much VRAM does SEX_ROLEPLAY-3.2-1B need?

I1-IQ1_S is exactly 479,751,104 bytes (0.45 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is SEX_ROLEPLAY-3.2-1B's KV cache?

1.00 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.

Novaciano · text

SEX_ROLEPLAY-3.2-1B

Novaciano/SEX_ROLEPLAY-3.2-1B

SEX_ROLEPLAY-3.2-1B at I1-IQ1_S is exactly 479,751,104 bytes (0.45 GiB / 0.48 GB) — an effective 2.561 bits per weight, not the nominal 1. Its KV cache at 32K is 1.00 GiB.

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

Parameters

1.5B

Architecture

llama

16 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
I1-IQ1_S	0.45 GiB	479,751,104	2.561	—	mradermacher
I1-IQ1_M	0.47 GiB	499,805,120	2.668	—	mradermacher
I1-IQ2_XXS	0.50 GiB	533,228,480	2.847	—	mradermacher
I1-IQ2_XS	0.52 GiB	562,064,320	3.001	—	mradermacher
I1-IQ2_S	0.56 GiB	601,587,456	3.212	—	mradermacher
I1-IQ2_M	0.59 GiB	628,326,144	3.354	—	mradermacher
I1-Q2_K_S	0.60 GiB	640,860,160	3.421	—	mradermacher
I1-Q2_K	0.62 GiB	667,074,560	3.561	—	mradermacher
I1-IQ3_XXS	0.63 GiB	674,987,776	3.603	—	mradermacher
I1-IQ3_XS	0.68 GiB	733,991,744	3.918	—	mradermacher
I1-Q3_K_S	0.70 GiB	754,570,048	4.028	—	mradermacher
I1-IQ3_S	0.70 GiB	756,798,272	4.040	—	mradermacher
I1-IQ3_M	0.72 GiB	770,167,616	4.112	—	mradermacher
I1-Q3_K_M	0.75 GiB	803,722,048	4.291	—	mradermacher
I1-Q3_K_L	0.79 GiB	845,402,944	4.513	—	mradermacher
I1-IQ4_XS	0.82 GiB	882,697,856	4.712	—	mradermacher
I1-Q4_0	0.86 GiB	920,790,912	4.916	—	mradermacher
I1-IQ4_NL	0.86 GiB	920,790,912	4.916	—	mradermacher
I1-Q4_K_S	0.86 GiB	923,412,352	4.930	—	mradermacher
I1-Q4_K_M	0.89 GiB	955,459,456	5.101	—	mradermacher
I1-Q4_1	0.93 GiB	995,928,448	5.317	—	mradermacher
I1-Q5_K_S	1.00 GiB	1,073,163,136	5.729	—	mradermacher
I1-Q5_K_M	1.02 GiB	1,092,103,040	5.830	—	mradermacher
I1-Q6_K	1.15 GiB	1,237,286,848	6.606	—	mradermacher

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	0.13 GiB	0.13 GiB	—	16 / 0 / 0
8,192	0.25 GiB	0.25 GiB	—	16 / 0 / 0
16,384	0.50 GiB	0.50 GiB	—	16 / 0 / 0
32,768	1.00 GiB	1.00 GiB	—	16 / 0 / 0
65,536	2.00 GiB	2.00 GiB	—	16 / 0 / 0
131,072	4.00 GiB	4.00 GiB	—	16 / 0 / 0

Compare with

same modality, comparable size

vs Qwen3-4B vs HyperCLOVAX-SEED-Text-Instruct-1.5B vs Llama-3.2-1B-Instruct vs ced-base vs gemma-3-1b-it vs Qwen3-1.7B All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts I1-IQ1_S at roughly 0.79 GiB. The real file is 0.45 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

128,260

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does SEX_ROLEPLAY-3.2-1B need?: I1-IQ1_S is exactly 479,751,104 bytes (0.45 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is SEX_ROLEPLAY-3.2-1B's KV cache?: 1.00 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 SEX_ROLEPLAY-3.2-1B 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.