CORE-Fast
Lightweight BugTraceAI model for rapid triage, concise payload review, scanner output analysis, and short-form engineering guidance.
BugTraceAI
Models built for BugTraceAI, an agentic web pentesting framework.
BugTraceAI combines security-focused AI models and tooling for authorized application security research, finding triage, technical reporting, and remediation support across web pentesting workflows.
CORE-Pro
Higher-capacity BugTraceAI model for long-context investigation, professional reporting, exploit-chain review, and remediation planning.
CORE-Ultra (27B)
The most capable CORE model. 27B SFT-trained on bug bounty reports, malware research, and WAF evasion data. Optimized for structured PoC generation, tool calling, and agentic pipelines. Available in Q4_K_S (15 GB) and Q6_K (21 GB).
Apex (26B)
The Apex Predator of offensive security reasoning. An uncensored 26B MoE running Opus-style thinking, fine-tuned over elite bug bounty data. (Includes Q4 and F16 Master).
Framework
BugTraceAI is an agentic web pentesting framework designed for authorized security operations, defensive research, and evidence-based analysis.
BugTraceAI systems are intended for authorized security work, defensive research, education, and engineering support. Outputs should be validated by humans before operational use.