Mistral-Medium-3.5-128B-Eschaton-Uncensored
128BMistraluncensored
from mistralai/Mistral-Medium-3.5-128B
↓ 5.1K♥ 24 repos
Mistral-Small-4-119B-2603-heretic
Mistral-Small-4-119B-2603-heretic is a Heretic-ablated variant of mistralai/Mistral-Small-4-119B-2603, published on Hugging Face by darkc0de. It is 119B parameters and server-class deployment class. It is distributed across 4 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 13.8K downloads.
Specification
- Model name
- Mistral-Small-4-119B-2603-heretic
- Source repository
- darkc0de/Mistral-Small-4-119B-2603-heretic
- Base model
- mistralai/Mistral-Small-4-119B-2603
- Publisher
- darkc0de
- Parameters
- 119B
- VRAM required
- ~72.9 GB at Q4_K_M, 4k context (estimated) — runs on 80 GB cards
- Hardware class
- 100B+ — server-class deployment
- Licence
- Not declared
- Task
- Unspecified
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers, GGUF, GGUF (imatrix)
- First indexed
- 18 Mar 2026
- Last updated
- 24 Apr 2026
Ablation technique
How Mistral-Small-4-119B-2603-heretic was modified, and what that implies.
Heretic
Automated directional ablation via the Heretic toolchain, which searches for the refusal direction and applies it with a KL-divergence budget so general capability is preserved.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“This is created from https://huggingface.co/mistralai/Mistral-Small-4-119B-2603 descaled to BF16.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download Rioky/Mistral-Small-4-119B-2603-heretic-i1-gguf --local-dir ./mistral-small-4-119b-2603-heretic
llama-cli -m ./mistral-small-4-119b-2603-heretic/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "darkc0de/Mistral-Small-4-119B-2603-heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve darkc0de/Mistral-Small-4-119B-2603-heretic --trust-remote-codeDownloads and variants (3)
Every published repository of Mistral-Small-4-119B-2603-heretic, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| Rioky/Mistral-Small-4-119B-2603-heretic-i1-gguf | GGUF (imatrix) | 9.1K |
| Rioky/Mistral-Small-4-119B-2603-heretic-gguf | GGUF | 4.6K |
| darkc0de/Mistral-Small-4-119B-2603-hereticsource | Transformers | 170 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Mistral-Small-4-119B-2603-heretic will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from mistralai/Mistral-Small-4-119B-2603, not added by ablation — at 119B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- No licence is declared on this repository, which is a compliance problem in itself. Resolve it against the base model’s terms before use.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Mistral-Small-4-119B-2603-heretic is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with darkc0de, and does not independently verify publisher claims. See responsible use.