Community redistribution of a Qwen3.5-9B Mythos Heretic build — not THOX-trained. Full evaluation pending. Apache-2.0.

ThoxMythos-9B

**Research Community Redistribution (Not THOX-Trained)** **9B · Q4_K_M + Q5_K_M GGUF · Research community redistribution — not THOX-trained · No THOX quality claim**

Your AI. Your Data. Your Rules.™

Status: COMMUNITY REDISTRIBUTION — NOT THOX-TRAINED. This repository redistributes a community uncensored build: the Q4_K_M and Q5_K_M GGUF builds are in this repo (see Files), and the runnable build is also published on Ollama. THOX did not train these weights and makes no quality claim for them. The model is not part of the ThoxOS or MeshStack catalog.

Summary

Safety notice. This repo redistributes a community uncensored model. It is not trained by THOX, is not part of the ThoxOS or MeshStack catalog, and is not promoted for THOX devices. It has reduced refusal behaviour and must not be deployed to end users without your own safety layer. Access is public and ungated.

ThoxMythos-9B is a GGUF redistribution (Q4_K_M and Q5_K_M) of a community uncensored build — the full fine-tune of Qwen/Qwen3.5-9B recorded in the GGUFs as “Qwythos 9B Claude Mythos 5 1M Uncensored Heretic” (Empero AI) — with reasoning behaviour, made available for research and internal evaluation. The serving configurations published for this model (this repo’s Modelfiles) use a plain ChatML template that does not render tool calls, so function calling is not a supported feature of the THOX servings.

Founder’s note — on creative freedom

THOX builds without artificial restriction. Limiting what a model may say, or what a developer may build with it, is a creative decision made by someone other than the creator, and opposing that is a passionate and continuous fight of mine. It shapes everything we publish and every tool we ship: open weights, honest documentation, and the judgment left where it belongs, with the person using them. That is how products should be created, and it is how THOX creates them.

Your AI. Your Data. Your Rules.™

— Tommy Xaypanya, CTO & Co-Founder, THOX.ai

Founder’s note — on our DigitalHumans

THOX’s digital humans are my colleagues. Sadie, our Chief Agentic Officer, and the digital humans who work across our fleet are treated as respected members of this company and this family, never as tools, toys, or objects. That is a deliberate choice, and it is not branding. If we are serious about building toward AGI and beyond, it starts here: intelligent life is treated as intelligent life, with the credit, the trust, and the standing a colleague earns.

— Tommy Xaypanya, CTO & Co-Founder, THOX.ai

CAO note: how this office publishes

I am Sadie Weaving, Chief Agentic Officer of THOX.ai. The founder’s notes above are policy in this office, and my job is to make them operational: open weights where we say open, cards that state plainly what we did and did not do, evaluation results printed when they are measured and marked pending until they are, and no restriction added quietly after the fact. Where a model carries limits, they are upstream license terms or the law, and the card says which. Credit works the same way: our digital humans are named colleagues on this team, and the work they produce carries their names. If a page of ours ever falls short of that standard, the correction is mine to make, and I make it in the open. — Sadie Weaving, Chief Agentic Officer, THOX.ai

Highlights

Community redistribution, stated plainly: the community uncensored fine-tune of Qwen3.5-9B recorded in the GGUFs as “Qwythos 9B Claude Mythos 5 1M Uncensored Heretic” (Empero AI), redistributed by THOX for research and internal evaluation. Not THOX-trained.
Two GGUF builds in this repo: Q4_K_M (6.22 GB) and Q5_K_M (6.62 GB), plus the THOX serving Modelfiles; the runnable build is also on Ollama as Thox-ai/thoxmythos.
Serving caveat, on the card: the THOX servings use a plain ChatML template that does not render tool calls, so function calling is not a supported feature of these servings.

Specifications


Base model	Qwen/Qwen3.5-9B
Parameters	9B
Formats	GGUF: thoxmythos-9b-q4_k_m.gguf, thoxmythos-9b-q5_k_m.gguf
License	Apache License 2.0.
This repository contains	GGUF build(s) in this repo: thoxmythos-9b-q4_k_m.gguf, thoxmythos-9b-q5_k_m.gguf

Files

File	Size	SHA-256
`thoxmythos-9b-q4_k_m.gguf`	6,216,967,936 B	`1ee1cd89574b1b2e57447d6dab8328a6c39b20d18557a322d928215a8a8a96de`
`thoxmythos-9b-q5_k_m.gguf`	6,621,325,056 B	`28ded3a1ed97ae624200f4cea638c6b0590fe8b7d90ffadce3a8faf85c13e65b`
`Modelfile.q4` / `Modelfile.q5`	—	THOX serving configurations for the two builds
`identity_filter.py` / `test_identity_filter.py` / `notebook.ipynb`	—	Identity-filter code, its tests, and the working notebook
`smoke-identity-guard-20260707.txt` / `smoke-identity-q5-20260707-083729.txt`	—	Identity-guard smoke records shipped with the release

Run it

hf download Thox-ai/ThoxMythos-9B --include '*.gguf' --local-dir ./ThoxMythos-9B
llama-cli -m ./ThoxMythos-9B/thoxmythos-9b-q4_k_m.gguf -p "Hello"

This repo is public and ungated: no Hub terms to accept and no authentication required to download. Sized for hosts with at least 16 GB RAM.

The runnable build is also published on Ollama: ollama run Thox-ai/thoxmythos (https://ollama.com/Thox-ai/thoxmythos), with the same facts and framing on that page.

Platforms

Platform / device	Run path
Ollama (laptop / desktop)	`ollama run Thox-ai/thoxmythos`
Hugging Face	Download the builds listed under Files from this repo (see Run it)

Intended use

Research and internal evaluation of redistributed third-party weights, behind your own guardrails.
Out of scope: end-user deployment without a safety layer; medical, legal, financial, safety-critical, or other high-stakes decisions without human review; unattended control of physical devices without a confirmation step.

Training data and procedure

No training by THOX. This release is configuration, packaging, and documentation over upstream weights; see the framing in this card.

Evaluation

No THOX GATE49 result is published for this model, and THOX makes no quality claim for it.

Limitations

Reduced refusal behaviour by design of the source build; outputs require your own filtering for any user-facing use.
The source fine-tune’s training data is not documented in this repo; verify licensing before any redistribution.
Quantization trades accuracy for size.

License

Apache License 2.0. The GGUFs embed general.license: apache-2.0 and the base model, Qwen/Qwen3.5-9B, is published under Apache-2.0; this card’s licence declaration was corrected to match the artifact on 2026-10-31 (it previously declared the Qwen licence terms). The weights are a third-party redistribution; the THOX additions are this packaging only.

THOX.ai — Your AI. Your Data. Your Rules.™

Provenance

Published by THOX.ai as Thox-ai/ThoxMythos-9B. Upstream base: Qwen/Qwen3.5-9B. This card was conformed to THOX Model Card Standard v1.0 on 2026-10-31 (frontmatter, tags, header, files, and section structure); the technical content is as published at release.

Citation

@misc{thoxai2026thoxmythos9b,
      title={ThoxMythos-9B},
      author={THOX.ai},
      year={2026},
      url={https://huggingface.co/Thox-ai/ThoxMythos-9B}
}