Llama

Skills for Meta's open-source Llama models. Best for self-hosting, multimodal AI, privacy-focused applications, and cost-effective deployment at scale.

1038 skills available

Meta’s Llama is the most widely deployed open-source AI model family. Llama 4 introduced mixture-of-experts architecture and native multimodal capabilities, with Scout offering an unprecedented 10M token context window. Fully open-weight and free to use, Llama powers everything from personal projects to national AI initiatives.

Key Strengths:

  • Fully open-weight with no API costs or vendor lock-in
  • 10M token context window (Scout) — process massive codebases and archives
  • Native multimodal understanding of text, images, and video
  • Trained on 200+ languages for global applications
  • Fine-tunable for domain-specific needs
  • Self-hostable for complete data control
  • Adopted by nations for sovereign AI initiatives (France, India, UAE)

Best For: Self-hosted deployments, privacy-sensitive applications, custom fine-tuning, multilingual projects, enterprise AI at scale, and developers building AI-powered products.

Featured Articles

View all

Browse All Skills

Showing 1038 of 1038 skills available
Advanced 20 min

Build Retrieval-Augmented Generation systems that ground LLM responses in external knowledge sources. Reduce hallucinations and enable …

4,500 4.7
Beginner 2 min

Comprehensive research, analysis, and content extraction system. Multi-source parallel research with deep content analysis for thorough …

1,200 4.6
Intermediate 3 min

Master utility-first CSS with TailwindCSS. Responsive design, dark mode, custom themes, and component patterns for rapid UI development.

391 4.7
Intermediate 5 min

Build accessible UI components with shadcn/ui. Beautifully designed components built on Radix UI and styled with Tailwind CSS.

428 4.8

Can't find what you need?

Tell us what AI skill or prompt you're looking for and we'll create it for you.

Request a Skill