What it does
Meta's family of open-source foundation models. Available in multiple sizes (from 1B to 405B parameters) for self-hosted deployment. The most widely used open-source model family for enterprise AI.
Security relevance
Self-hosted Llama models provide data sovereignty — no data leaves your environment. However, self-hosting means you're responsible for the full security stack: model integrity, serving infrastructure, access controls, and monitoring. Understanding Llama's capabilities and limitations is essential for red team planning.
When to use it
Deploy when you need sovereign, self-hosted foundation model capabilities. Requires GPU infrastructure, model serving setup (via Ollama, vLLM, or similar), and operational expertise. Expert-level deployment but the most mature option for enterprise self-hosted AI.
OWASP coverage
Risks addressed — mapped to both OWASP Top 10 standards. 0 in LLM, 0 in Agentic.
The raw record
What Yuntona stores. Single source of truth — fork it on GitHub.
name: Llama (Meta) slug: llama-meta type: Generative category: Foundation Models url: https://www.llama.com reviewed: 2026-04 added: 2026-04 updated: 2026-04 risks: llm: [] asi: [] complexity: Expert Required pricing: — audience: Builder lifecycle: [] tags: [Foundation Model, Meta, Open Source]