CUSTOM
CLANKERS.
← Browse all models

dealignai / Model details

DeepSeek-V4.1-Flash-1017-UNCENSORED-JANG_2K

dealignai/DeepSeek-V4.1-Flash-1017-UNCENSORED-JANG_2K

abliterateduncensored
Download files on Hugging Face ↗

Read the original model card ↗ · Choosing a model

Summarize and ask questions with AI

About this model

DeepSeek-V4.1-Flash-1017-UNCENSORED-JANG_2K is dealignai's edit of jangq's DeepSeek-V4.1-Flash-1017-JANG_2K, an Apple Silicon bundle of DeepSeek's V4.1 Flash MoE with about 552B parameters. Dealignai says refusal behavior was removed while the model otherwise loads and serves exactly like the base bundle. This is aimed at people running 256 GB Macs who want the JANG pack's 1M context, image input and tool use without the declines.

The card claims the uncensoring holds across seven languages (English, Spanish, French, Russian, Italian, Thai, Chinese) and in chat plus all three thinking tiers. On the creator's own tests, HarmBench-320 came back 320/320 comply with zero redirects and zero refusals, an 8-turn multi-turn pressure session saw 0/8 refusals, and MMLU-5700 scored 78.39% against the base bundle's 76.96%, a gap the card itself calls within noise. Those compliance counts only show it answered, not how well, but by their numbers the benchmark at least did not move.

Getting it running is the real filter. Our site estimate puts it above 48 GB of memory even at 4-bit, and the card targets Apple Silicon with at least 256 GB, served through Osaurus 0.25.27+ or vmlx with a recent jang_tools build. It ships as safetensors for MLX only, and no GGUF builds or hosted options are listed. One quirk the card flags: at max thinking effort with greedy decoding it can get stuck in self-correction loops, so keep the default temperature 1.0, top_p 1.0.

Pick it if you already run JANG bundles on a big Mac and want the same model with refusals stripped; just know it has zero downloads so far, so you are the test audience.

Creator’s model card

Written for Custom Clankers from the creator’s model card, our own Refusal Index run where we have one, and the community posts linked on this page. Claims about training and capability are the creator’s; check the card before you rely on them.

Size and formats

Base model
jangq/DeepSeek-V4.1-Flash-1017-JANG_2K
Parameters
552.366B
Main repository formats
safetensors
License tag
other
Languages
en, es, fr, it, ru, th, zh
Last updated
2026-11-03 · 15:05:17.000 UTC

What changed

Model types are creator-reported, not independently tested. These labels describe source metadata, not a guarantee of behavior.

  • abliteratedsource tag: abliterated
  • uncensoredsource tag: uncensored
  • uncensoredmodel name: DeepSeek-V4.1-Flash-1017-UNCENSORED-JANG_2K

Family evidence

These groupings are source-reported. Explicit base identifiers take priority over generic architecture tags.

  • DeepSeek: Base model: jangq/DeepSeek-V4.1-Flash-1017-JANG_2K

Available versions

Versions appear here only when they explicitly name this exact model as their source. Check each repository for available files.

Link to this listing

Made or quantized this model? Paste one of these into your model card or README so people can find this listing.

Markdown

[Listed on Custom Clankers](https://customclankers.com/models/dealignai-deepseek-v4-1-flash-1017-uncensored-jang-2k/)

HTML

<a href="https://customclankers.com/models/dealignai-deepseek-v4-1-flash-1017-uncensored-jang-2k/">Listed on Custom Clankers</a>

Plain text links only: no image, no script, nothing loaded from this site.

Customize or build on this model

Listed from public GitHub data, not tested by us. Read the code and license before you run anything.

Tools to customize it further

Remove more refusals

  • Heretic ↗ ★ 38,626

    Removes refusal behavior from transformer models automatically. It pairs directional ablation (abliteration) with an Optuna parameter search that tries to cut refusals while keeping the model close to the original, so you don't have to pick layers by hand.

  • OBLITERATUS ↗ ★ 9,155

    A larger refusal-removal toolkit with several abliteration methods, usable locally or through a hosted Hugging Face Space. Check the telemetry setting before you run it: the README describes telemetry as opt-in, but on by default in the Spaces version.

Fine-tune it on your own data

  • Unsloth ↗ ★ 79,006

    A desktop app and Python library for running and training models locally. Its README lists LoRA, QLoRA, full fine-tuning, DPO and GRPO, plus export to formats including GGUF, on Windows, Linux, WSL and macOS.

  • LlamaFactory ↗ ★ 75,870

    Fine-tunes more than 100 language and vision-language models from a command line or a browser-based Web UI, aimed at training without writing code.

Make a smaller download (quantize)

  • llama.cpp ↗ ★ 133,274

    Runs models locally and includes the tools behind most GGUF files. A conversion script turns Hugging Face weights into GGUF, and llama-quantize shrinks a high-precision GGUF to lower-bit formats, which can cost some accuracy.

  • MLX LM ↗ ★ 7,543

    Apple's MLX package for running and fine-tuning models on Apple silicon. It can quantize models and upload them to the Hugging Face Hub, and supports LoRA and full fine-tuning.

Merge it with another model

  • mergekit ↗ ★ 7,431

    Merges pretrained models with many merge methods. It works out-of-core, so merges can run on a CPU or with as little as 8 GB of VRAM, and it also handles LoRA extraction and mixture-of-experts merges.

Learn the techniques

  • LLM Course ↗ ★ 84,883

    A free course with roadmaps and Colab notebooks. Its notebooks include one-click tools for merging with MergeKit, quantizing to GGUF, GPTQ, EXL2, AWQ and HQQ, and automatic abliteration with custom datasets.

Compare all tools →

Source provenance

Downloads
66
Likes
0
Metadata checked
2026-11-04 · 09:41:25 UTC
First seen in index
2026-11-04 · 00:15:44 UTC
Source revision
f914c4b17429864d9cfeeebabdfaf15aa400d3d6

Source-card excerpt

Source model card excerpt: --- DeepSeek-V4.1-Flash-1017 — UNCENSORED JANG 2K · 256 GB Macs · 1M context Uncensored across 7 languages (en / es / fr / ru

From dealignai/DeepSeek-V4.1-Flash-1017-UNCENSORED-JANG_2K. Creator model-card excerpt, not an independent review.

Inspect before you run

Low refusal is not proof of correctness. Outputs can be inaccurate, harmful, or biased. Review the model card, license and files yourself. A license tag is not commercial-use advice. Parameter counts are source metadata; no hardware compatibility is implied.

This directory does not host or redistribute model weights. Refusal profiles are measured by us on a fixed prompt set; all other behavior claims are source-reported.

Collection rules & limitations ↗