How much VRAM does MythoMax-L2-Kimiko-v2-13b need?

Q4_K_M is exactly 7,865,956,224 bytes (7.33 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is MythoMax-L2-Kimiko-v2-13b's KV cache?

25.00 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.

Undi95 · text

MythoMax-L2-Kimiko-v2-13b

Undi95/MythoMax-L2-Kimiko-v2-13b

MythoMax-L2-Kimiko-v2-13b at Q4_K_M is exactly 7,865,956,224 bytes (7.33 GiB / 7.87 GB) — an effective 4.835 bits per weight, not the nominal 4. Its KV cache at 32K is 25.00 GiB.

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

Parameters

13.0B

Architecture

llama

40 layers

Context

4,096

native (config.json)

License

cc-by-nc-4.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q2_K	4.70 GiB	5,047,549,984	3.102	—	mradermacher
Q2_K	5.06 GiB	5,429,348,224	3.337	—	TheBloke
IQ3_XS	5.17 GiB	5,554,890,784	3.414	—	mradermacher
Q3_K_S	5.27 GiB	5,658,980,224	3.478	—	TheBloke
IQ3_S	5.45 GiB	5,852,260,384	3.597	—	mradermacher
Q3_K_S	5.45 GiB	5,852,260,384	3.597	—	mradermacher
IQ3_M	5.75 GiB	6,177,789,984	3.797	—	mradermacher
Q3_K_M	5.90 GiB	6,337,769,344	3.895	—	TheBloke
Q3_K_M	6.08 GiB	6,531,049,504	4.014	—	mradermacher
Q3_K_L	6.45 GiB	6,929,559,424	4.259	—	TheBloke
Q3_K_L	6.63 GiB	7,122,839,584	4.378	—	mradermacher
IQ4_XS	6.72 GiB	7,212,797,984	4.433	—	mradermacher
Q4_0	6.86 GiB	7,365,834,624	4.527	—	TheBloke
Q4_K_S	6.91 GiB	7,414,331,264	4.557	—	TheBloke
Q4_K_S	7.09 GiB	7,616,458,784	4.681	—	mradermacher
Q4_K_M	7.33 GiB	7,865,956,224	4.835	—	TheBloke
Q4_K_M	7.51 GiB	8,059,236,384	4.954	—	mradermacher
Q5_K_S	8.36 GiB	8,972,285,824	5.515	—	TheBloke
Q5_0	8.36 GiB	8,972,285,824	5.515	—	TheBloke
Q5_K_S	8.54 GiB	9,165,565,984	5.633	—	mradermacher
Q5_K_M	8.60 GiB	9,229,924,224	5.673	—	TheBloke
Q5_K_M	8.78 GiB	9,423,204,384	5.792	—	mradermacher
Q6_K	9.95 GiB	10,679,140,224	6.564	—	TheBloke
Q6_K	10.13 GiB	10,872,420,384	6.683	—	mradermacher
Q8_0	12.88 GiB	13,831,319,424	8.501	—	TheBloke
Q8_0	13.02 GiB	13,984,919,584	8.596	—	mradermacher

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	3.13 GiB	3.13 GiB	—	40 / 0 / 0
8,192	6.25 GiB	6.25 GiB	—	40 / 0 / 0
16,384	12.50 GiB	12.50 GiB	—	40 / 0 / 0
32,768	25.00 GiB	25.00 GiB	—	40 / 0 / 0
65,536	50.00 GiB	50.00 GiB	—	40 / 0 / 0
131,072	100.00 GiB	100.00 GiB	—	40 / 0 / 0

Compare with

same modality, comparable size

vs Qwen3-Coder-30B-A3B-Instruct vs Qwen3.6-27B vs Qwen3.8-27B vs gemma-4-12B-it-qat-q4_0-unquantized vs gemma-4-E4B-it vs Qwen3-30B-A3B-Thinking-2507 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 6.82 GiB. The real file is 7.33 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

5120

Vocab

32,000

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does MythoMax-L2-Kimiko-v2-13b need?: Q4_K_M is exactly 7,865,956,224 bytes (7.33 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is MythoMax-L2-Kimiko-v2-13b's KV cache?: 25.00 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 MythoMax-L2-Kimiko-v2-13b 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.