# Llms

> Large language models basics

- Skill: `ffsshhttiikk/llms` (Agent Skill)
- Install (CLI): `npx skillmds@latest add ffsshhttiikk/llms`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ffsshhttiikk/llms/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: ffsshhttiikk (https://skillmd.com/u/ffsshhttiikk)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/ffsshhttiikk/llms

---


## What I do

- Apply LLMs to solve problems
- Design LLM applications
- Create effective prompts
- Evaluate LLM outputs
- Optimize LLM usage

## When to use me

Use me when:
- Building AI-powered products
- Automating text tasks
- Need natural language understanding

## Key Concepts

### LLM Capabilities
- **Text generation**: Complete, continue, summarize
- **Question answering**: Extract, generate
- **Translation**: Multilingual
- **Classification**: Sentiment, intent
- **Code**: Generation, completion
- **Reasoning**: Chain-of-thought

### Model Selection
- **GPT-4**: General purpose, strong reasoning
- **Claude**: Helpful, Claude constitutional AI
- **LLaMA**: Open source, efficient
- **Mistral**: Fast, open source
- **Code models**: CodeGen, CodeLlama

### Cost Optimization
- Use smaller models when possible
- Implement caching
- Batch requests
- Optimize prompt length
- Use streaming for long outputs