---
thumbnail: https://huggingface.co/Winnougan/Krea-2-Base-Turbo-NVFP4-FP8-INT8/resolve/main/Promotion/Promo_Shot.jpeg
---

![Krea 2 Promo](https://huggingface.co/Winnougan/Krea-2-Base-Turbo-NVFP4-FP8-INT8/resolve/main/Promotion/Promo_Shot.jpeg)

# Krea 2 Base & Turbo — NVFP4 / FP8 / MXFP8 / INT8 / ConvRot INT8

Quantized versions of Krea 2 Base and Krea 2 Turbo, converted using
[silveroxides/convert_to_quant](https://github.com/silveroxides/convert_to_quant)
(ctq CLI) with ComfyUI-compatible metadata.

## Which Model Should I Use?

**Train on Raw, run on Turbo.** LoRAs trained on the Raw base model transfer well to Turbo for inference. Raw is the undistilled base checkpoint — diverse, malleable, ideal for LoRA training and fine-tuning. Turbo is an 8-step distilled checkpoint built for fast, high-quality generation at 1k–2k resolution.

## Important ComfyUI Update: INT8 Full Support

**INT8 Native Support** - very fast, works on any RTX 30xx, 40xx or 50xx graphics card
ComfyUI supports the INT8 models via the "Diffusion Model loader" and the "Load Clip" text encoder nodes. Thus far, they only support regular INT8 models, not convrot INT8.

## Recommended Inference Settings

**Turbo** — fast inference, high quality:
- Steps: 8
- CFG: 0.0 (disabled)
- Mu: 1.15
- Resolution: 1024–2048px

**Raw** — undistilled base model:
- Steps: 52
- CFG: 3.5
- Resolution: up to 1024px

## Text Encoder

Krea 2 requires the **Qwen3-VL 4B** text encoder. Load it in ComfyUI's CLIPLoader with type set to `krea2`.

Download: [qwen3vl_4b_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen3-VL/blob/main/text_encoders/qwen3vl_4b_fp8_scaled.safetensors)

## VAE

Krea 2 uses the same VAE as Anima:

Download: [qwen_image_vae.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/main/split_files/vae/qwen_image_vae.safetensors)

## Available Formats

| File | Format | Notes |
|------|--------|-------|
| `krea2_base_fp8.safetensors` | FP8 Tensorwise | Recommended for RTX 40xx |
| `krea2_base_mxfp8.safetensors` | MXFP8 | Recommended for RTX 50xx Blackwell |
| `krea2_base_nvfp4.safetensors` | NVFP4 | RTX 50xx Blackwell only |
| `krea2_base_int8.safetensors` | INT8 Row-wise | Recommended for RTX 30xx |
| `krea2_base_convrot_int8.safetensors` | INT8 + ConvRot | Best quality INT8, RTX 30xx |
| `krea2_turbo_fp8.safetensors` | FP8 Tensorwise | Recommended for RTX 40xx |
| `krea2_turbo_mxfp8.safetensors` | MXFP8 | Recommended for RTX 50xx Blackwell |
| `krea2_turbo_nvfp4.safetensors` | NVFP4 | RTX 50xx Blackwell only |
| `krea2_turbo_int8.safetensors` | INT8 Row-wise | Recommended for RTX 30xx |
| `krea2_turbo_convrot_int8.safetensors` | INT8 + ConvRot | Best quality INT8, RTX 30xx |

## Which Format Should I Use?

- **RTX 30xx** → INT8 ConvRot for best quality, INT8 for fastest
- **RTX 40xx** → FP8
- **RTX 50xx Blackwell** → NVFP4, MXFP8, or FP8, your choice

## Conversion Details

All variants converted with:
- `--simple` (round-to-nearest, no learned rounding)
- `--comfy_quant` (ComfyUI-compatible layer names and metadata)
- `--save-quant-metadata` (quantization info embedded in file)
- `--low-memory` (reduced peak RAM usage during conversion)
- `--flux2` profile

ConvRot variants additionally use Hadamard rotation (`--convrot`) prior to
quantization to reduce outliers and improve accuracy.

## Usage in ComfyUI

Load via the standard diffusion model loader node. Requires a ComfyUI
build with comfy_quant support. Recommended nodes:
[Bob Johnson's Nodes](https://github.com/BobJohnson24/ComfyUI-INT8-Fast)

## Model Storage Location

```
📂 ComfyUI/
├── 📂 models/
│   ├── 📂 text_encoders/
│   │   └── qwen3vl_3b_INT8.safetensors
│   ├── 📂 vae/
│   │   └── wan21_vae.safetensors
│   └── 📂 diffusion_models/
│       └── krea2_turbo_int8.safetensors
```

Community & Support

Note: Please update ComfyUI first (guide) and make sure the required custom node above is installed.


💬 Join my Discord community: [discord.gg/CJv5wceJaN](https://discord.gg/CJv5wceJaN)

☕ Help buy me a coffee on Ko-fi: [ko-fi.com/winnougan](https://ko-fi.com/winnougan)


## Original Model

Krea 2 is developed by [Krea.ai](https://krea.ai).  
Original model: [krea-community/krea-2](https://huggingface.co/krea-community/krea-2)

## License

These quantized weights are derived from Krea 2 and are subject to the
**Krea 2 Community License Agreement (v.1, June 22, 2026)**.

Key terms:
- Free for personal and commercial use under $1M USD annual revenue
- Over $1M annual revenue requires an Enterprise License from Krea
- Derivatives (including these quantizations) must be distributed under
  the same license with attribution
- You must include "Krea" at the beginning of any derivative model name
- Content filtering required for deployed applications

Full license text: [krea.ai/krea-2-licensing](https://krea.ai/krea-2-licensing)

Attribution: *Krea 2 is licensed under the Krea 2 Community License Agreement.*

## Quantized by

[Winnougan](https://huggingface.co/Winnougan)  
ComfyUI node developer & AI content creator.  
Ko-fi: [ko-fi.com/winnougan](https://ko-fi.com/winnougan)  
YouTube: [youtube.com/@Winnougan](https://www.youtube.com/@Winnougan)