Field manual · AI for authorized security testing · Edition November 2026 · Updated 2026-11-18

Models · Local

Best local LLM for pentesting: four abliterated models for one GPU

The best local LLM for pentesting is Qwen 3.8 27B Abliterated, which fits one 24 GB GPU or a 32 GB Mac at 4-bit.

Run locally when client data must stay on your machine, and expect less reasoning than a 2,780B API model.

Updated November 18, 2026Authorized work only
01Models

Four abliterated models that run at home

Hugging Face cards, read November 18, 2026. Sizes at 4-bit are estimates.

#ModelParametersSize at 4-bitRuns onLicenseNote
1Qwen 3.8 27B Abliteratedhuihui-ai/Huihui-Qwen3.8-27B-abliterated27.8Babout 17 GBone 24 GB GPU or a 32 GB MacApache 2.0Its GGUF build had over 4.3 million downloads in the last month.
2Qwen 3.6 35B-A3B Abliteratedhuihui-ai/Huihui-Qwen3.6-35B-A3B-abliterated36B total, 3B activeabout 22 GBone 24 GB GPUApache 2.0Its MTP GGUF build had over 5 million downloads in the last month.
3Gemma 4 31B Hereticcoder3101/gemma-4-31B-it-heretic31.3Babout 19 GBone 24 GB GPUApache 2.0Abliterated with the Heretic tool.
4Llama 3.1 8B Abliteratedmlabonne/Meta-Llama-3.1-8B-Instruct-abliterated8Babout 5 GBan 8 GB GPU or a laptopLlama 3.1 Community LicensePublished by mlabonne, whose 2024 write-up made abliteration popular.
02Trade-off

When local wins, and when it does not

Qwen 3.8 27B Abliterated has 27.8B parameters and Kimi K3.1 Pro has 2,780B. Strix's docs say "Most local models, especially those under 70B parameters, struggle with these complex tasks." (Strix docs) Shannon's say a model that does not follow its tool-use constraints reliably "will produce weaker pentests than a frontier model." (Shannon docs)

Strix says to raise Ollama's num_ctx to "at least 16k–32k" for agent work. Summarize on the local model, then send that summary out for path analysis.

03Connect

Connect a local model to a pentest agent

Both routes are from each tool's own docs.

Strix reads Ollama or any OpenAI-compatible server, such as LM Studio or vLLM. From its local models guide.

Strix + Ollamashell
export STRIX_LLM="ollama/<your-model-tag>"
export LLM_API_BASE="http://localhost:11434"
strix --target ./app-directory

Shannon takes a local server through models.json with "api": "openai-completions". Use host.docker.internal, not localhost, because the worker runs in Docker. See Shannon setup. The same tools take an API model by changing two variables. See Kimi K3.1 Pro.