Model

LLM is the category. A model is a specific instance.

Anthropic has several Claude models, each with different trade-offs:

• Larger models are smarter but slower and more expensive.
• Smaller models are faster and cheaper, often good enough.
• Newer models generally beat older ones — but not always.

Models are versioned like software. Claude Opus 4.7. Claude Sonnet 4.6. Claude Haiku 4.5. When you read API docs or tutorials, the model name is what you pass to specify which brain you want to talk to.

Mosaic's familiar uses Claude. Which specific model changes over time as new ones release — check app/api/familiar/route.ts for the current one.