Skip to content
Back to Glossary
◉ ai

Open Weights

AI models anyone can use and modify

What is Open Weights?

AI models whose trained parameters (weights) are publicly released for anyone to download, run, fine-tune, and build upon. Llama, Mistral, and DeepSeek are prominent open-weight families. Unlike proprietary APIs, open weights let you run models locally, customize them fully, and avoid vendor lock-in. For Web3 builders who value decentralization, open-weight models align with the ethos of user-owned infrastructure.

Take the next useful step

LV’s AI hub

Original context on AI applied to creation, research, and systems.

Creator OS

Useful when AI is part of a content workflow that still needs human judgment.

💡

In plain words

"Think of it like a recipe published online — you can tweak it, remix it, or run it as-is on your own stove."

How it works

Key takeaways

  • Download and run on your own hardware

  • Full control over data privacy and customization

  • Open weights ≠ open source (training code may be closed)

▸

Real-world example

Meta releases Llama weights: anyone can download them and run the model on their own hardware. No API dependency, no vendor lock-in. You control the model, the data, and the infrastructure entirely.

Related terms

LLM (Large Language Model)

AI models that understand and generate text

Fine-Tuning

Training an AI model on your specific data

Inference

Using a trained AI model to generate outputs

Copied