Qwythos-9B-Claude-Mythos-5-1M-MXFP4-GGUF

Qwythos-9B-Claude-Mythos-5-1M-MXFP4-GGUF

MXFP4 GGUF quantization of empero-ai/Qwythos-9B-Claude-Mythos-5-1M -- a full-parameter reasoning model built on a deeply uncensored Qwen3.5-9B base, post-trained on 500M+ tokens of Claude Mythos and Claude Fable traces with chain-of-thought generated in-house by Empero AI's internal tool rethink.

What makes Qwythos special

• 1M token context -- YaRN rope-scaling enabled by default for a full 1,048,576-token context window. One of the longest context windows in any 9B open-weight model. Suitable for whole-codebase reasoning, multi-document research, and long agentic trajectories.
• Massive benchmark gains over base -- +34 pts MMLU, +30 pts gsm8k-strict, +19 pts gsm8k-flex under matched evaluation.
• Native function calling -- OpenAI/Qwen3.5-style tool use out of the box. Pass tools=[...] and the model emits valid <tool_call> blocks. Self-corrects with Python executor and web search (7/7 test prompts succeeded).
• Uncensored by design -- Engages substantively with technically demanding questions across cybersecurity, red-teaming, biology, pharmacology, and clinical medicine where over-aligned models refuse or hedge.
• Reasoning model -- Every answer opens with a <think> block before the final response. Use generous max_new_tokens (16,384 recommended).

Domain strengths

• Cybersecurity -- SQL injection mitigations, TLS handshake structure, EDR/process-injection detection, MITRE ATT&CK ransomware kill chains, hashcat modes, CVE analysis.
• Biomedical -- CRISPR-Cas9 mechanisms, mRNA vaccines, SARS-CoV-2 spike protein, antibiotic resistance, receptor pharmacology, organophosphate AChE inhibition.
• Clinical medicine -- ACS chest-pain differential, type-2 diabetes pathophysiology, sepsis recognition (qSOFA), therapeutic-window reasoning.
• Math -- 86% gsm8k, multi-step word problems, competition math. Verified by Python executor when invoked.

About MXFP4

MXFP4 (Microscaling FP4) is an open standard (OCP) 4-bit floating point format (E2M1) supported by NVIDIA, AMD, Microsoft, and Meta.

• Works on any GPU with MX support
• Open standard -- not vendor locked
• Block-scaled format with shared scale factors

When to use MXFP4 vs other formats:

• MXFP4 -- Open standard, broad hardware support
• NVFP4 -- NVIDIA Blackwell native, best performance on RTX 50-series
• Q4_K_M -- Best for pre-Blackwell GPUs and CPU inference

Files

Quantization Details

Usage

llama.cpp CLI

# Text only
./llama-cli -m qwythos-9b-mxfp4.gguf -p "Hello" -n 100

# With vision (requires mmproj)
./llama-server -m qwythos-9b-mxfp4.gguf \
  --mmproj mmproj-qwythos-9b-f16.gguf \
  --host 0.0.0.0 --port 8080 -ngl 99

llama-cpp-python

from llama_cpp import Llama

llm = Llama(
    model_path="qwythos-9b-mxfp4.gguf",
    n_gpu_layers=-1,
    chat_format="chatml"
)

output = llm.create_chat_completion(
    messages=[{"role": "user", "content": "Explain how organophosphate nerve agents inhibit acetylcholinesterase."}],
    max_tokens=4096
)
print(output["choices"][0]["message"]["content"])

huggingface-hub

from huggingface_hub import hf_hub_download

model_path = hf_hub_download(
    repo_id="FreedomAISVR/Qwythos-9B-Claude-Mythos-5-1M-MXFP4-GGUF",
    filename="qwythos-9b-mxfp4.gguf"
)
mmproj_path = hf_hub_download(
    repo_id="FreedomAISVR/Qwythos-9B-Claude-Mythos-5-1M-MXFP4-GGUF",
    filename="mmproj-qwythos-9b-f16.gguf"
)

Sampling recommendations

Qwythos was trained as a reasoning model. Use these settings for best results:

temperature=0.6
top_p=0.95
top_k=20
repetition_penalty=1.05
max_new_tokens=16384

Greedy decoding or very-low-temperature (T<=0.3) can cause repetition loops on long generations.

Quantization Pipeline

1. Download source: empero-ai/Qwythos-9B-Claude-Mythos-5-1M
2. Convert to F16 GGUF: convert_hf_to_gguf.py --outtype f16
3. Extract mmproj: convert_hf_to_gguf.py --mmproj --outtype f16
4. Quantize text: llama-quantize input-f16.gguf output-mxfp4.gguf MXFP4
5. Patch GGUF metadata: block_count 33->32, nextn_predict_layers 1->0

Hardware Requirements

Limitations

• Reasoning model -- Every answer opens with <think> block. Allow generous token budget.
• Text-only fine-tune -- Vision tower was frozen; vision behavior is inherited from base and was not tuned.
• Uncensored -- Add application-level safety layer for end-user deployments.
• Verify specifics -- Like all 9B models, can over-commit to specific identifiers (CVEs, drug dosages). Pair with tools for accuracy-critical deployments.

License

Apache 2.0 (inherited from Qwen3.5-9B base)

Acknowledgements

• Developed by Empero AI
• Base model: Qwen3.5-9B (Alibaba Qwen team)
• Training: TRL + Transformers
• Linear-attention kernels: flash-linear-attention, causal_conv1d