Get the app
← back to catalog · registered 2026-08-22 13:56

Uniboshi/Kimi-K3-Abliterated-V1-GGUF

Uniboshi Kimi GGUF multimodal second-order 1.0M ctx
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/Uniboshi%2FKimi-K3-Abliterated-V1-GGUF"
Response includes
  • classification m8
  • files 3
  • hub_downloads_all_time 1,071
  • author_summary 2 models
  • readme_text full
Try next
GET/models?author=Uniboshi
All 2 models by Uniboshi
 GET/models?family=kimi&method=m8
Sibling kimi models · same method
 GET/methods/m8
What "m8" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
M1 Direct removal →
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=1
  • assume M1 (base ablation) + M8 (GGUF quant) - default when producer unknown
How we label → About M8 →
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
165 last 30d - stable
Likes
35
Model age
2mo ago
created 2026-08-02

Training datasets

Corpora the author lists in the model card. Datasets tracked in our /datasets catalog carry a category badge linking to the workflow stage. Others open on Hugging Face.

TFMC/imatrix-dataset-for-japanese-llm
Downloads over time
Now1.1K→from238↑355%
1965198431.2K238 on Aug 51.1K on Oct 111.1K on Oct 7AugSepOct
Aug 5 → Oct 11 · 50 snapshots · spans 67 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.

Uniboshi/Kimi-K3-Abliterated-V1↓ 1,988 this lineage
Uniboshi/Kimi-K3-Abliterated-V1-GGUF↓ 165

Variants by this author 2 formats · 2K downloads combined

The same weights this author released in different packaging. Pick the format that matches your runtime.

Metadata

License
other
Tags
transformers gguf conversational abliterated uncensored image-text-to-text dataset:TFMC/imatrix-dataset-for-japanese-llm base_model:Uniboshi/Kimi-K3-Abliterated-V1 base_model:quantized:Uniboshi/Kimi-K3-Abliterated-V1 license:other endpoints_compatible region:us

Related

Other by Uniboshi
Kimi-K3-Abliterated-V1 2K
Same base model
GrEarl/Kimi-K3-Abliterated-V1-Q2_K-GGUF 99
Total size
3.18 GB
Files
3
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-08-13 05:51
Source huggingface.co/Uniboshi/Kimi-K3-Abliterated-V1-GGUF ↗ Base model Uniboshi/Kimi-K3-Abliterated-V1 →

Files by quantization

Auxiliary files 3 files 3.18 GB
imatrix.gguf 3.18 GB ******** download
.gitattributes 5.59 KB 1f490f68 download
README.md 937 B beda1c39 download

README current version from Hugging Face

base_model:

  • Uniboshi/Kimi-K3-Abliterated-V1
    base_model_relation: quantized
    tags:
  • gguf
  • conversational
  • abliterated
  • uncensored
    license: other
    license_name: "kimi-k3"
    library_name: transformers
    pipeline_tag: image-text-to-text
    datasets:
  • TFMC/imatrix-dataset-for-japanese-llm

Uniboshi/Kimi-K3-Abliterated-V1-GGUF

  • These models are GGUF-converted versions of Uniboshi/Kimi-K3-Abliterated-V1.
  • The code and chat templates used for model conversion are from Unsloth.
  • For model quantization, the dataset used as calibration data to calculate the importance matrix is ​​TFMC/imatrix-dataset-for-japanese-llm, which consists solely of English and Japanese data.
  • By the way, calculating the importance matrix was more cost-intensive than creating the Abliterated model. LMAO!

Discussions 1 thread

  1. 2026-08-05Severe repetition and degraded output with Kimi-K3-Abliterated IQ1_Sopen2 💬#1
    Loading...

Continue on Hugging Face →

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.

Get the app Browse the catalog
Open in Abliteration