Launching today

PentestLLM
Uncensored AI for hacking and pentesting, one API key
0 followers
Uncensored AI for hacking and pentesting, one API key
0 followers
Security-tuned, abliterated and uncensored models on one prepaid API key for authorized pentesting, red teaming, bug bounty and CTF work. Works in Claude Code, OpenCode and Cursor, with free CVSS and rules-of-engagement tools. Card from $10, no subscription.



Launch Team

AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted
Hi Product Hunt. PentestLLM is the security-side door to the model catalog we run: the security-tuned, abliterated and uncensored builds (Gemma, Qwen, GLM, DeepSeek and the OrcaCyber line) on one prepaid API key, for people doing authorized pentesting, red teaming, bug bounty and CTF work.
Why it exists: the hosted assistants refuse a large share of ordinary security prompts (exploit analysis, payload review, writing a PoC against your own lab). An uncensored build answers them, and a model built for security work answers them better. The key works in Claude Code, OpenCode and Cursor through the OpenAI and Anthropic formats, and in PyRIT and garak for red-teaming AI systems you are authorized to test.
The site also has free tools: a CVSS calculator, a rules-of-engagement builder and an engagement cost estimator. Every model gets a public daily check that the id answers as the model named. Card from $10, no subscription.
What we sell is authorized work only, and the acceptable-use page says so. Happy to answer questions on which builds we serve and how the daily check works.