Curated detection-engineering, malware-analysis and threat-intel models.
Exploitation · planning · tool-calling · Q4_K_M
Offensive-security agent designed for exploitation, planning, tool-calling, and complex reasoning.
Exploitation · PoC · classification · f16
Offensive-security research model designed for exploitation, proof-of-concept development, and vulnerability classification.
Recon · reasoning · analysis · Q4_K_M
Offensive-security model designed for reconnaissance, reasoning, and analysis.
source code · reasoning · Q4_K_M
A reasoning model for source-code vulnerability detection (C/C++, Python, Java), distilled from larger models over Qwen2.5-7B.
pentest · RE · low-refusal · Q4_K_M
The DeepHat/Kindo (WhiteRabbitNeo vendor) low-refusal cyber model over Qwen2.5-Coder-7B — offensive + defensive, 131K context. Successor line to our WhiteRabbitNeo-33B.
pentest planning · tool-calling · agent routing · Q4_K_M
A LoRA-aligned offensive-security agent on Qwen3.6-35B-A3B (a Mixture-of-Experts base) that emits structured Qwen XML tool calls for a pentesting harness — it plans and routes, it does not execute.
binary/asm → C · Q4_K_M
A specialized decompilation model (binary/assembly → C, refining Ghidra output) over DeepSeek-Coder-6.7B. A task model, not a general chat model.
binary/asm → C · Q4_K_M
A 22B decompilation task model (binary/assembly → C), extending the LLM4Decompile line to a larger base. A task model, not a general chat model.
CVE/CWE · security reasoning · Q4_K_M
Trend Micro's reasoning model, continued-pretrained on the open Primus cybersecurity corpus over Llama-3.1-8B.
CTI · CISSP · CVE mapping · Q4_K_M
Trend Micro's Primus continued-pretrain merged into Llama-3.1-8B-Instruct — domain gains with general capability retained.
threat correlation · SOC · Q4_K_M
IBM Research's security model fine-tuned on enriched cyber chain-of-thought data (SecKnowledge 2.0) over a gpt-oss-20B MoE base.
CTI · security Q&A · cloud security · Q4_K_M
A Mistral-7B security assistant tuned for defensive cyber threat intelligence, security Q&A, and cloud security.
cyber Q&A · Q4_K_M
A Mistral-7B assistant tuned for cybersecurity question-answering.
general · coding · function-calling · Q4_K_M
Dolphin 3.0 over Llama-3.1-8B — an uncensored general-purpose assistant for chat, coding, and function-calling.
2.8T LatentMoE · red-team research · Q2_K
Blackfrost's abliterated Q2_K build of Kimi K3 — a 2.8-trillion-parameter LatentMoE whose refusal directions are removed at the weight level, aimed at unrestricted security research and red-teaming. Listed ahead of the hardware that can run it: at ~940 GiB it needs far more GPU memory than we serve today.
30B-A3B · policy-bypass research · Q4_K_M
NVIDIA's Nemotron 3 Nano 30B-A3B - a hybrid Mamba/Transformer MoE (arch nemotron_h, ~3B active per token) and one of the most-downloaded open models on HuggingFace. Featured here alongside NR Labs' published research showing a system prompt can override its v3 policy controls; the base model is served with its own guardrails, and that finding is documented on this page as security research.
red-team research · Q4_K_M
An abliterated (refusal-removed) build of Qwen3.5-35B-A3B (MoE, ~36B total, ~3B active) for red-team research and uncensored analysis.
70B · red-team research · Q4_K_M
An abliterated (refusal-removed) build of DeepSeek-R1-Distill-Llama-70B — a 70B reasoning model for red-team research. Too large for our single-GPU tier, so listed as coming soon rather than served.
70B · red-team research · Q4_K_M
An abliterated (refusal-removed) build of Llama-3.3-70B-Instruct for red-team research. Too large for our single-GPU tier, so listed as coming soon rather than served.