Qwen · image

Qwen-Image-2512

Qwen/Qwen-Image-2512

Qwen-Image-2512 at Q4_K_M is exactly 13,065,746,976 bytes (12.17 GiB / 13.07 GB) — an effective 5.116 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)
Parameters
20.4B
Architecture
qwen_image
Context
—
native (config.json)
License
apache-2.0

Shipped quantizations

● exact bytes, summed from published files
QuantSize●Exact bytes●Effective bpw●Tensors●Publisher
Q2_K6.83 GiB7,333,837,3442.872—unsloth
Q3_K_S8.59 GiB9,223,928,3523.612—unsloth
Q3_K_M9.25 GiB9,932,896,8003.8891933unsloth
Q4_011.04 GiB11,852,773,9204.6411933unsloth
Q4_K_S11.43 GiB12,268,010,0164.804—unsloth
Q4_111.96 GiB12,843,678,2405.029—unsloth
Q4_K_M12.17 GiB13,065,746,9765.1161933wikeeyang
Q4_K_M12.34 GiB13,244,758,5605.1861933unsloth
Q5_K_S13.32 GiB14,298,184,2245.599—unsloth
Q5_013.41 GiB14,400,813,6005.639—unsloth
Q5_K_M13.97 GiB15,000,074,7845.8741933unsloth
Q5_114.33 GiB15,391,717,9206.027—unsloth
Q6_K15.67 GiB16,824,990,2406.5881933unsloth
Q8_020.27 GiB21,761,817,1208.5211933unsloth
Q8_020.27 GiB21,761,817,1208.5211933wikeeyang
BF1638.07 GiB40,872,114,72016.004—unsloth
F1638.07 GiB40,872,114,72016.0041933unsloth

Pipeline components

a diffusion model is a graph of parts, not one file
ComponentSizeShareCan live on the CPU?
denoiser38.05 GiB71%no, must be resident
vae0.24 GiB0%no, must be resident
text encoder15.45 GiB29%yes — the usual first VRAM saving
Full pipeline53.74 GiBresident if nothing is offloaded

The parameter count published for a diffusion model describes the denoiser alone. Running it also requires its text encoder and VAE, and the text encoder is often nearly as large as the denoiser — which is why offloading it is the standard first move when you run out of memory.

We publish component sizes here, not throughput. Community-submitted image-generation rates do exist for many GPUs and we show them on the hardware pages, but they aggregate runs at different resolutions, step counts and settings, so they cannot be attributed to one model. Peak memory during sampling is unmeasured by any public source, and we do not estimate it.

Compare with

same modality, comparable size
vs FLUX.2-klein-9Bvs Qwen-Image-Edit-2511vs FLUX.1-devvs Krea-2-Turbovs FLUX.2-klein-4Bvs Z-Image-TurboAll 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 10.70 GiB. The real file is 12.17 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

Architecture unavailable — this repository is gated and no ungated mirror was found. Exact file sizes above are still authoritative; only the KV math needs the config.

Questions people ask

How much VRAM does Qwen-Image-2512 need?
Q4_K_M is exactly 13,065,746,976 bytes (12.17 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of Qwen-Image-2512 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.