Open Source / Open Weights

Models or tools whose source code or weights are publicly available to use, study, and self-host.

"Open weights" (e.g. Llama, Qwen, Mistral) means you can download and run the model yourself, which is key for privacy and offline use — though licenses vary and aren't always fully open-source.

Related resources

Deploying Local LLMs with Ollama

A cross-platform tool that lets you download and run open-source LLMs like Llama, Mistral, and Phi locally on your own machine. It provides a CLI, a local API server, and an OpenAI-compatible endpoint, so you can experiment with models privately — no GPU or cloud account required. Ideal for developers who want offline AI capabilities or need to test prompts before deploying to production.

PyTorch Deep Learning Framework

PyTorch is a leading open-source deep learning framework developed by Meta, known for its dynamic computation graph and Pythonic interface. It powers both cutting-edge AI research and production deployments, with a rich ecosystem of tools for training, debugging, and scaling models across GPUs and TPUs.

Llama is Meta's family of open-weight language models — the backbone of the open-source AI ecosystem, deployable locally or on any cloud. Available in multiple sizes from lightweight to frontier-class, Llama models power countless fine-tuned derivatives and give organizations full control over their AI stack.