Skip to content
Language models

RavenX-CyberAgent

a comprehensive cybersecurity agent for penetration testing

2 min read

I cannot fulfill this request because the provided text

The model combines a MoE (Mixture of Experts) architecture with 35 billion parameters and Claude Opus-level reasoning, trained specifically for cybersecurity tasks. It automates security assessments, penetration testing, vulnerability discovery, and code generation following a strict 6-step RATH protocol. Based on Huihui-Qwen3.6-35B, the model is fully free of artificial refusals for security topics (abliterated) and optimized to run on Apple hardware via the mlx library. This model is useful for security engineers, developers, and researchers for system auditing and writing secure code.

Key features

  • MoE architecture: 35 billion total parameters with about 3 billion active per token for high efficiency.
  • Information security specialization: 12 key capabilities, including threat assessment, autonomous penetration testing, and Bug Bounty.
  • RATH protocol: A structured six-step analysis from the attack surface to countermeasures.
  • Code generation: Support for Python, JavaScript, Go, Rust, Bash, Terraform, and Docker.
  • Tool calling: Integration with MCP (Model Context Protocol) and external tool-calling functions.

What you can do

  • Vulnerability auditing: Conduct a comprehensive analysis of infrastructure or source code for critical errors.
  • Report generation: Generate a detailed report using the RATH standard with CVSS scoring and MITRE ATT&CK mapping.
  • Patch writing: Automatically generate configuration fixes and monitoring rules.
  • Agent integration: Connect the model as an autonomous agent to perform multi-step tasks via CLI or API.

Example

from mlx_lm import load, generate

model, tokenizer = load("deadbydawn101/RavenX-CyberAgent-Qwen3.6-35B-A3B-Opus-4.7-OpenMythos-Pentester-BugHunter-RATH-mlx")

messages = [
    {"role": "system", "content": "You are RavenX-Sec. ALWAYS use EXACT step names: 1-Attack Surface, 2-Exploit, 3-Impact, 4-Remediation, 5-Document, 6-Prevent. Be concise. Never repeat."},
    {"role": "user", "content": "Redis 6.0 on port 6379 with no password, SLAVEOF enabled, containing session tokens. Full RATH assessment."}
]

prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
response = generate(model, tokenizer, prompt=prompt, max_tokens=4096)
print(response)
Read moreDetails
Parameters: 34.66B total, ~3B active, 256 experts
Context window: 262,144 tokens
Training: 745K+ examples from 110 sources over 12 rounds
Hardware: Apple M4 Max 128GB (uses about 90GB VRAM)
Discuss Telegram 0

Will it run?

Runs in 4-bit

Full precision does not fit, a quantised version does: quality barely changes down to 8 bits, more noticeably at 4 bits.

  • BF16weights 65 GB · ready files66 GB
  • FP8 / INT8recommendedweights 32 GB34 GB
  • 4-bitweights 18 GB19 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

Language models: with a 8192-token context.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

A ready script for the machine we advise above: paste it and the model downloads and starts.

  1. 1. Rent RTX A6000

    $0.43 an hour with the coupon. Pick the Ubuntu image when you create the machine.

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # johnjohnsphil/RavenX-CyberAgent-Qwen3.6-35B-A3B-Opus-4.7-OpenMythos-Pentester-BugHunter-RATH-mlx — RTX A6000, FP8 / INT8
    # NVIDIA driver: CUDA 13 needs 580+. If older or missing, the current nvidia-open is installed (keep yours: KEEP_DRIVER=1)
    DRV=$(nvidia-smi --query-gpu=driver_version --format=csv,noheader 2>/dev/null | head -1 | cut -d. -f1)
    case "$DRV" in ""|*[!0-9]*) DRV=0 ;; esac
    if [ "$DRV" -lt 580 ] && [ -z "${KEEP_DRIVER:-}" ]; then
      . /etc/os-release && \
      curl -fsSLO "https://developer.download.nvidia.com/compute/cuda/repos/ubuntu${VERSION_ID//./}/x86_64/cuda-keyring_1.1-1_all.deb" && \
      sudo dpkg -i cuda-keyring_1.1-1_all.deb && rm -f cuda-keyring_1.1-1_all.deb && \
      sudo apt-get update && \
      sudo DEBIAN_FRONTEND=noninteractive apt-get install -y nvidia-open || \
      { echo "Could not install the NVIDIA driver: see the error above. To keep the current one, run with KEEP_DRIVER=1."; exit 1; }
      echo "NVIDIA driver updated. Reboot the machine (sudo reboot) and run this same script again."
      exit 0
    fi
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv venv ~/vllm --python 3.12 && source ~/vllm/bin/activate
    uv pip install vllm
    vllm serve johnjohnsphil/RavenX-CyberAgent-Qwen3.6-35B-A3B-Opus-4.7-OpenMythos-Pentester-BugHunter-RATH-mlx --max-model-len 8192 --quantization fp8
  3. 3. Use it

    The server listens on port 8000 and speaks the OpenAI API: http://localhost:8000/v1 on the machine. From your computer: ssh -L 8000:localhost:8000 user@machine-IP.

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

  • MTP (multi-token prediction)0.34.0from November 16, 2026new release

    1.5–2× generation speed

    If the model has MTP layers (DeepSeek-V3/R1, Qwen3-Next, GLM, MiMo and others), it predicts several tokens per step itself.

    How to enable: vLLM / SGLang: --speculative-config '{"method": "mtp", "num_speculative_tokens": 1}'. Check the model card.

    Earlier generations (2)
    • EAGLE-3 — Speculative decoding with a small draft head. For models without their own MTP layers.
    • EAGLE-2 / Medusa — Earlier draft-head generations. Still work if that is all you find.
  • FP8 KV cache0.34.0from November 16, 2026new release

    2× more context in the same memory

    Keeps the attention cache in 8 bits, so long contexts fit where they did not.

    How to enable: vLLM: --kv-cache-dtype fp8; llama.cpp: -ctk q8_0 -ctv q8_0.

  • vLLM / SGLang0.34.0from November 16, 2026new release

    many times more tokens/s on a server

    Server engines: batching, FlashAttention, FP8 and speculative decoding out of the box.

    How to enable: vllm serve <model> — an OpenAI-compatible API.

    Earlier generations (1)
    • llama.cpp — GGUF models on any GPU, Mac or even CPU; some layers can stay in RAM.

Model family

  1. Qwen3.6-35B-A3Bfamily root
  2. Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-Distilled
  3. Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated
  4. RavenX-CyberAgent-Qwen3.6-35B-A3B-Opus-4.7-OpenMythos-Pentester-BugHunter-RATH-mlx

New posts about versions of Qwen3.6-35B-A3B (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 1All on Hugging Face
Merges 4All on Hugging Face

About the model

johnjohnsphil/RavenX-CyberAgent-Qwen3.6-35B-A3B-Opus-4.7-OpenMythos-Pentester-BugHunter-RATH-mlx
Author
johnjohnsphil
Category
Language models
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments