← Back to the A-Z glossary
AI glossary · P2

Open source vs open weight

Open weight means the model's weights are available. Open source also implies more transparency about the code, and sometimes the data.

Concept 4 min read Updated 2026-05-22
— Definition

Open source vs open weight, in plain words

Open weight means the model's weights are available. Open source also implies more transparency about the code, and sometimes the data.

Clear up a very common misuse of the terms.

A concrete example

A model can be downloadable without its training data or its full code being public.

Why it matters

The difference changes what you can audit, modify, host, or use commercially.

You'll see it in discussions around Llama, Mistral, Qwen, Hugging Face, and local models.

Don't mix it up with

LLM: An LLM is an AI model trained to understand and generate language.

Fine-tuning: Fine-tuning adapts an existing model on examples to improve one specific behavior.

Common mistakes

  • Saying open source the moment a model is downloadable.
  • Ignoring the license.
  • Forgetting the commercial-use restrictions.

Quick checklist

  • First I check whether the word names a concept, a tool, a risk, or a metric.
  • I tie it to a concrete case: A model can be downloadable without its training data or its full code being public.
  • I keep the main trap in mind: Saying open source the moment a model is downloadable.

Quick questions

What is Open source vs open weight in AI?

Open weight means the model's weights are available. Open source also implies more transparency about the code, and sometimes the data.

Where will I run into Open source vs open weight?

You'll see it in discussions around Llama, Mistral, Qwen, Hugging Face, and local models.

Which word should I read next?

Start with LLM, Fine-tuning, Quantization.

Want to keep going in order?

Head back to the full glossary, search a word, then open only the pages that deserve more than a short definition.

Open the AI glossary