Qwen · text

QwQ-32B

Qwen/QwQ-32B

QwQ-32B at Q4_K_M is exactly 19,851,335,840 bytes (18.49 GiB / 19.85 GB) — an effective 4.847 bits per weight, not the nominal 4. Its KV cache at 32K is 8.00 GiB.

From the file· summed from 1 file(s)From the file· KV per layer
Parameters
32.8B
Architecture
llama
64 layers
Context
40,960
native (config.json)
License
—

Shipped quantizations

● exact bytes, summed from published files
QuantSize●Exact bytes●Effective bpw●Tensors●Publisher
UD-IQ1_S7.16 GiB7,685,786,2081.877—unsloth
UD-IQ1_M7.73 GiB8,298,465,8882.026—unsloth
UD-IQ2_XXS8.70 GiB9,336,801,8882.280—unsloth
UD-IQ2_M10.71 GiB11,495,189,0882.807—unsloth
Q2_K11.47 GiB12,313,098,8483.006—unsloth
Q2_K_L11.64 GiB12,495,575,6483.051—unsloth
UD-IQ3_XXS12.16 GiB13,052,652,1283.187—unsloth
Q3_K_S13.40 GiB14,392,330,8483.514—unsloth
Q3_K_M14.84 GiB15,935,048,2883.891771unsloth
Q3_K_L16.06 GiB17,247,078,9764.211—lmstudio-community
IQ4_XS16.50 GiB17,717,484,1284.326771unsloth
IQ4_NL17.40 GiB18,682,174,0484.562—unsloth
Q4_017.43 GiB18,711,009,8884.569771unsloth
Q4_K_M18.49 GiB19,851,335,8404.847771Qwen
Q4_K_M18.49 GiB19,851,336,2564.847—lmstudio-community
Q4_K_M18.49 GiB19,851,336,2884.847—unsloth
Q4_119.22 GiB20,639,242,8485.040—unsloth
Q5_021.08 GiB22,638,254,2405.528—Qwen
Q5_K_M21.66 GiB23,262,156,9605.680771Qwen
Q5_K_M21.66 GiB23,262,157,4085.680771unsloth
Q6_K25.04 GiB26,886,154,4006.565771Qwen
Q6_K25.04 GiB26,886,154,8166.565771lmstudio-community
Q6_K25.04 GiB26,886,154,8486.565—unsloth
Q8_032.43 GiB34,820,884,6408.502771Qwen
Q8_032.43 GiB34,820,884,6728.502—unsloth
Q8_032.43 GiB34,820,885,0568.502771lmstudio-community
BF164 shards122.07 GiB131,071,939,55232.004—unsloth

KV cache by context

computed per layer
ContextKV cache (f16)●Flat formulaOverstated byFull / windowed / recurrent
4,0961.00 GiB1.00 GiB—64 / 0 / 0
8,1922.00 GiB2.00 GiB—64 / 0 / 0
16,3844.00 GiB4.00 GiB—64 / 0 / 0
32,7688.00 GiB8.00 GiB—64 / 0 / 0
65,53616.00 GiB16.00 GiB—64 / 0 / 0
131,07232.00 GiB32.00 GiB—64 / 0 / 0

Compare with

same modality, comparable size
vs Qwen3-Coder-30B-A3B-Instructvs Qwen3.6-27Bvs Qwen3.8-27Bvs gemma-4-12B-it-qat-q4_0-unquantizedvs gemma-4-E4B-itvs Qwen3-30B-A3B-Thinking-2507All alternatives →Will it fit on my card? →

Will it run on your card?

full quant x context sweep
By VRAM:8GB12GB16GB24GB32GB48GB
Radeon RX 6500 XT 4GBGeForce RTX 3050 6GBGeForce RTX 5050 8GBGeForce RTX 3080 10GBGeForce RTX 2080 Ti 11GBGeForce RTX 5070 12GBGeForce RTX 5060 Ti 16GBApple M3 Pro 18GBGeForce RTX 3080 Ti 20GBGeForce RTX 5090 D V2 24GBGeForce RTX 5090 D 32GBApple M5 Max 36GBApple M5 Pro 48GBApple M5 Max 64GBApple M3 Ultra 96GBApple M5 Max 128GBApple M2 Ultra 192GBApple M3 Ultra 256GBApple M3 Ultra 512GB

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 17.16 GiB. The real file is 18.49 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json
Layers
64
Attention heads
40
KV heads
8
Head dim
128
Hidden size
5120
Vocab
152,064
Sliding window
32768
SWA period
—
MLA
no
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 QwQ-32B need?
Q4_K_M is exactly 19,851,335,840 bytes (18.49 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is QwQ-32B's KV cache?
8.00 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 QwQ-32B 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.