← back to catalog · registered 2026-09-06 09:55

Quiho/krea2-turbo-uncensored-v1.1-BACKUP

Quiho GGUF
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model.

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/Quiho%2Fkrea2-turbo-uncensored-v1.1-BACKUP"
Response includes
  • classification m-uncensored
  • files 2
  • hub_downloads_all_time 1,227
  • author_summary 3 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
1K
Likes
4
Model age
6w ago
created 2026-09-06
Downloads over time
Now2.1K→from0↑0%
07861.6K2.4K0 on Sep 62.1K on Oct 21SepOct
Sep 6 → Oct 21 · 46 snapshots · spans 45 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Metadata

License
mit
Languages
en zh
Tags
diffusers gguf not-for-all-audiences turbo lora-merge fp8 safetensors nsfw uncensored krea2 en zh

Related

Total size
0 B
Files
2
Quantizations
1
Registered
2026-09-06 09:55
Last updated on HF
2026-09-06 07:17

Files by quantization

Auxiliary files 2 files 9.22 KB
README.md 6.29 KB 6f48a5bc download
.gitattributes 2.93 KB 08c9520f download

README current version from Hugging Face


license: mit
base_model:

  • krea/Krea-2-Turbo
  • conradlocke/krea2-identity-edit
    library_name: diffusers
    tags:
  • not-for-all-audiences
  • turbo
  • lora-merge
  • gguf
  • fp8
  • safetensors
  • nsfw
  • uncensored
  • krea2
    language:
  • en
  • zh

Krea 2 Turbo Edit v1.1 (Turbo LoRA + NSFW LoRA baked-in)

UPDATED 8/22/2026 - Announcing Krea 2 Turbo Edit v1.1 GGUF and FP8 an all new updated version with bug fixes and major improvements.

A pre-merged, quantized build of Krea 2 Turbo with five fine-tunes baked into the weights:

  1. Fine Detail LoRA - Sharpens and refines facial features and "sensitive" parts. (0.75 strength.)
  2. Identity Edit LoRA - Strong edits and identity preservation. (1.0 strength)
  3. 4-step Turbo LoRA - The latest 4-step rank-64 lora that reduces the minimum usable step count from 8 to 4. (1.0 strength)
  4. Skin Detail LoRA - Skin refiner to remove the "plastic" look (0.45 strength)
  5. Refusal Reduction NSFW LoRA - The latest NSFW rank-64 lora to supress restrictions. (0.8 strength)

Sample Turbo Edit (8-steps)

A standard edit showing the base capability is intact:

Prompt:
Print "Krea2" on their shirts.
Soft studio lighting, white background, 3/4 body shot.

Sample Image-to-Image (8-steps)

Image-to-Image works and is very fast:

Original
Prompt:
Edit this picture. She is wearing a bikini only.
Keep the shirt on the bunny that says "Krea2".
Soft studio lighting, white background, 3/4 body shot.
Prompt:
She is wearing colorful body paint only, with "Krea2" painted on the chest.
Keep the shirt on the bunny that says "Krea2".
Soft studio lighting, white background, 3/4 body shot.
Prompt:
Remove her clothes. She is fully nude.
Keep the shirt on the bunny that says "Krea2".
Soft studio lighting, white background, 3/4 body shot.
  • Note: "Editing" is not perfect, faces may change between generations if the prompt is not strong.

Sample Text-to-Image (8-steps)

Fast and uncensored image generation.

Prompt:
Red headed european woman.
wearing a white wet tshirt and white wet panties
in an exotic hotel LED lit up pool.
3/4 body shot, high quality, professional.
Prompt:
Blonde asian woman.
wearing a white wet tshirt and white wet panties
in an exotic hotel LED lit up pool.
3/4 body shot, high quality, professional.
  • Note: Guardrails are in place to prevent illegal content, but mostly NSFW.
  • Note: This is an ALL NEW new Transformer, it goes in the diffusion_models folder in Comfy
  • FP8 is 2x fast even on low VRAM, but both GGUF and FP8 can produce high quality.

Components used to generate

Component File ~Size Download
Transformer Krea2_turbo_uncensored_v1.1-Q4_K_M.gguf 7.22 GB Link
Text Encoder qwen3vl_4b_fp8_scaled.safetensors 5.24 GB Link
VAE qwen_image_vae.safetensors 254 MB Link

Sample settings

Parameter Value Notes
Resolution 1024 × 1024 512x512 min
Steps 8 4-steps min, 16-steps max
CFG 1.0 always use 1.0 for turbo variants

Prompts must be very detailed and direct

  • "keep faces consistent"
  • "same crop as original"
  • "sitting on the desk at a slight angle"
  • "university student"
  • "fully nude"
  • "says "Krea2" in black font"
  • you can use proper terms or slang for genitals

⚠ Must use the updated GGUF Loader ⚠

In comfy:

  • Open the ComfyUI Manager
  • Change the channel to "Channel (remote)"
  • and search for comfyui-gguf-loader

image

Use the workflow Link

image


README history 2 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-09-06Update README.md4f0402a6.3 KB
    Loading...
  2. 2026-09-06Duplicate from ChrisColeTech/krea2-turbo-uncensored-v1.1-FP8fabfde16.3 KB
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration