Skip to content

Stanford Alpaca

A compact instruction-following model from 52K demos

Visit official site

What Stanford Alpaca is

The model is described as a LLaMA 7B fine-tuned on 52K instruction demonstrations at Stanford CRFM.

What you can do with it

  • Follow instructions in one turn
  • Generate text from a given prompt
  • Answer questions put to the model

Model notes

  • A preliminary evaluation is mentioned
  • Single-turn behaviour is likened to text-davinci-003
  • Small scale and easy reproduction are stated

Pros & cons

✓ What we like

  • Reproduction is described as easy
  • The model size is stated as small

! What to watch out for

  • The evaluation is described as preliminary
  • Similarity is stated for single-turn instructions

FAQ

What is it based on?

The material states a LLaMA 7B model fine-tuned on 52K instruction demonstrations.

What was compared?

A preliminary evaluation compared it with text-davinci-003.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review