Guides
Which open-weight AI models run on which hardware, what's inside them, and how to pick one: guides built on the site's own measurements.
- The largest open-weight AI models, by parameter countThe biggest AI models anyone can download, ranked by total parameters, with their size on disk and how many parameters each uses per token. Updated automatically as new models come out.
- Best AI models for an 8 GB GPU (RTX 4060, RTX 5060)Which open-weight LLMs actually run on an 8 GB graphics card, which file to download, and how much context is left. Updated automatically as new models come out.
- Best AI models for a Mac with 64 GB of memoryWhich open-weight LLMs run on an Apple Silicon Mac with 64 GB, which file to use, and how much context fits. Updated automatically as new models come out.
- Best AI models for a 24 GB GPU (RTX 4090, RTX 3090)Every open-weight LLM that runs on a 24 GB graphics card, with the file to download and how much context fits next to it. Updated automatically as new models come out.