Free AI Tool Now Shows Which Models Fit Your Computer
You can finally filter AI models by your computer's memory without any tech skills.
Models can now report their trained context length without loading the model: a new helper opens only a GGUF's metadata and reads .context_length, accepting both u32 and u64 values and returning 0 when the file is missing, unreadable, invalid, or reports no context length.
On the server side, update_caps — which already resolves the model file offline to read its modalities — now also reads the trained context from the same GGUF metadata, and GET /models reports it as context_length when known. A router listing then carries the context without any Hub request at all, so the UI can sort and filter by it offline. In the UI, the option mapping fills contextLength from the router's report, and the manager reads it before the Hub record. The Context column, the context sort, and the context filter then work with the Hugging Face Hub API turned off — and a browser suite guards the sort and the search, the Hub-cache driven context filter, and the re-sort when details arrive after the sort was clicked.
Favorites also got attention: a favor
- The app now shows how much text each AI model can remember, so you can pick one that fits your computer.
- You can mark favorite models with a heart and find them instantly.
- Everything works offline, so no internet needed to sort and filter models.
Why It Matters
Saves time and prevents crashes by helping you choose the right AI model for your device.