eugeneyan/open-llms

open-llms is a curated reference list of large language models (LLMs) that are licensed for commercial use. Each entry in the list covers a specific model: when it was released, where to download the trained weights (called checkpoints), links to the research paper or announcement blog post, how many parameters the model has, how much text it can process at once (context length), and which license it uses. All models included have licenses that permit commercial use, such as Apache 2.0, MIT, or OpenRAIL-M. This makes the list practical for developers and businesses who want to build products with AI language capabilities but need to verify intellectual property terms first. The list covers a wide range, from small efficient models under one billion parameters to large ones such as Bloom at 176 billion parameters, spanning releases from 2019 through 2023. This is not software you run or a framework you install. It is a reference document maintained as a GitHub repository so anyone can contribute additions or corrections via pull requests. Each row in the main table links out to the actual model weights (hosted on sites like Hugging Face) and to the papers explaining how each model was built. The repository is primarily useful for teams evaluating which open, commercially licensed model to use in a product, or for researchers tracking what is publicly available. Community contributions are welcome. The full README is longer than what was shown.