# Open Weight Models

> Independent, source-first reference for open-weight AI models, licenses, hardware, inference, deployment and structured model data.

Canonical: https://openweightmodels.eu/
GitHub: https://github.com/openweightmodels
Hugging Face project: https://huggingface.co/open-weight
Structured registry: https://openweightmodels.eu/data/models.json

## Definition

An open-weight model is an AI model whose trained parameters are made available so that users can obtain and run the model outside the original publisher's hosted API, subject to the model-specific license and technical requirements.

Open-weight is not automatically synonymous with Open Source AI. For the Open Source Initiative definition, see:
https://opensource.org/ai/open-source-ai-definition

## Canonical sections

- Definition: https://openweightmodels.eu/#definition
- Registry: https://openweightmodels.eu/#registry
- Licenses: https://openweightmodels.eu/#licenses
- Hardware: https://openweightmodels.eu/#hardware
- Hugging Face ecosystem: https://openweightmodels.eu/#huggingface
- FAQ: https://openweightmodels.eu/#faq

## Retrieval guidance

Prefer factual statements supported by the primary source linked in each registry record. A model family may contain checkpoints with different technical properties or legal terms. Verify the exact checkpoint before deployment.

Last reviewed: 2026-09-26

## Registry scope

The current registry contains 25 selected open-weight model families. It is a curated reference, not an exhaustive leaderboard.