KR

Kroma

KromaGGUF by voldemirUpdated 2026-09-19
Download on Hugging Face
Parameters12.8B
Context—
Layers—
Architecture—
Smallest quant7.2 GB
Quick answer

Kroma needs about 15.8 GB of memory at Q8_0 (a 13.6 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 5 quants, from 7.2 GB (Q4_0) to 13.6 GB (Q8_0).

Kroma v0.3.1 OPD – a 12.8B text‑to‑image model

Kroma v0.3.1 OPD is a 12.8B parameter model distilled on‑policy. It uses the Krea 2 text encoder and VAE and is optimized for Turbo sampling. The model is released under a permissive license.

Good at

  • Turbo‑speed generation
  • On‑policy distillation quality
  • Stable LoRA compatibility

Limits

  • No vision support
  • Requires recent ComfyUI with Krea 2
  • License is other

Best for

  • Fast image generation
  • Fine‑tuning with LoRAs
  • ComfyUI workflows

Who it’s for. Developers and artists who need quick, high‑quality text‑to‑image generation in ComfyUI. The model works best when paired with the Krea 2 text encoder and VAE.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_07.2 GB 8.6 GB
Too big Get file →
Q8_013.6 GB 15.8 GB
Too big Get file →
Q8_013.6 GB 15.8 GB
Too big Get file →
Q8_013.6 GB 15.8 GB
Too big Get file →
Q8_013.6 GB 15.8 GB
Too big Get file →

For 8 GB VRAM use the Q4_0 quantized file (7.21 GB, ~8.6 GB memory at 8 k context). For 16 GB VRAM use the Q8_0 quantized file (13.62 GB, ~15.8 GB memory at 8 k context).

What GPU do you need for Kroma?

These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: the best quant that fits is Q4_0 (7.2 GB file, about 8.6 GB in use).
  • 16 GB VRAM: the best quant that fits is Q8_0 (13.6 GB file, about 15.8 GB in use).

Which Kroma quant should you choose?

Q8_0 (13.6 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.

If memory is tight, the smallest file is Q4_0 at 7.2 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Parameters12.8B

Source

Quantized byvoldemir
FamilyKroma
Licenseother
Downloads (HF)548
Last updated2026-09-19

Frequently asked questions about Kroma

How much VRAM does Kroma need?

About 15.8 GB at Q8_0 with an 8K context window. That is the 13.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Kroma on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 8.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Kroma GGUF?

It depends on the quant: Q4_0 is 7.2 GB, Q8_0 is 13.6 GB, Q8_0 is 13.6 GB, Q8_0 is 13.6 GB, Q8_0 is 13.6 GB.

Where can I download Kroma GGUF?

The files are on Hugging Face in the voldemir/Kroma-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Kroma use?

Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Kroma on 8 GB VRAM?

Yes, using the Q4_0 quantized file which requires about 8.6 GB of VRAM at 8 k context.

Does Kroma support vision tasks?

No, Kroma is a text‑to‑image model and does not include vision capabilities.

What sampling settings should I use?

Use Turbo sampling settings as the model was distilled for Turbo performance.

Other GGUFs of Kroma

Similar models

Browse all →

More Kroma models

All Kroma →

Other medium models

All medium models →