How much VRAM does polyglot-lion-0.6b need?

I1-IQ1_S is exactly 208,014,880 bytes (0.19 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

knoveleng · text

polyglot-lion-0.6b

knoveleng/polyglot-lion-0.6b

polyglot-lion-0.6b at I1-IQ1_S is exactly 208,014,880 bytes (0.19 GiB / 0.21 GB) — an effective 2.127 bits per weight, not the nominal 1.

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

Parameters

782M

Architecture

qwen3vl

null layers

Context

—

native (config.json)

License

mit

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
I1-IQ1_S	0.19 GiB	208,014,880	2.127	—	mradermacher
I1-IQ1_M	0.20 GiB	216,051,232	2.209	—	mradermacher
I1-IQ2_XXS	0.21 GiB	229,445,152	2.346	—	mradermacher
I1-IQ2_XS	0.23 GiB	241,995,296	2.474	—	mradermacher
I1-IQ2_S	0.24 GiB	254,193,184	2.599	—	mradermacher
I1-IQ2_M	0.25 GiB	264,908,320	2.709	—	mradermacher
I1-IQ3_XXS	0.26 GiB	279,014,944	2.853	—	mradermacher
I1-Q2_K_S	0.26 GiB	280,558,112	2.869	—	mradermacher
I1-Q2_K	0.28 GiB	296,237,600	3.029	—	mradermacher
I1-IQ3_XS	0.29 GiB	312,752,672	3.198	—	mradermacher
I1-Q3_K_S	0.30 GiB	323,074,592	3.303	—	mradermacher
I1-IQ3_S	0.30 GiB	323,074,592	3.303	—	mradermacher
I1-IQ3_M	0.31 GiB	336,026,144	3.436	—	mradermacher
I1-Q3_K_M	0.32 GiB	347,126,304	3.549	—	mradermacher
I1-IQ4_XS	0.34 GiB	367,802,912	3.761	—	mradermacher
I1-Q3_K_L	0.34 GiB	368,491,040	3.768	—	mradermacher
I1-IQ4_NL	0.36 GiB	381,565,472	3.901	—	mradermacher
I1-Q4_0	0.36 GiB	382,155,296	3.907	—	mradermacher
I1-Q4_K_S	0.36 GiB	383,269,408	3.919	—	mradermacher
I1-Q4_K_M	0.37 GiB	396,704,288	4.056	—	mradermacher
I1-Q4_1	0.38 GiB	409,090,592	4.183	—	mradermacher
I1-Q5_K_S	0.41 GiB	436,615,712	4.464	—	mradermacher
I1-Q5_K_M	0.41 GiB	444,414,496	4.544	—	mradermacher
I1-Q6_K	0.46 GiB	495,106,592	5.062	—	mradermacher

Compare with

same modality, comparable size

vs Llama-3.2-1B-Instruct vs ced-base vs embeddinggemma-300m vs Qwen3-0.6B vs LFM2.5-1.2B-Instruct vs MiniCPM5-1B-Claude-Opus-Fable5-Thinking 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 0.41 GiB. The real file is 0.19 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

—

Vocab

—

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does polyglot-lion-0.6b need?: I1-IQ1_S is exactly 208,014,880 bytes (0.19 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of polyglot-lion-0.6b 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.