Your first chat in the Hugging Bay app

1. Get the app for your computer

Follow the Mac or Linux installation steps, using the current files and checksums on our download page. Windows and Intel Mac downloads were not listed as available when checked on 17 October 2026; use Ollama or LM Studio where supported.

2. Let the app check your machine

Open the app and follow the first-run computer check. Choose Chat or Writing if asked what you want to do. Open Discover, select a small text chat model, and inspect the actual model file, license and memory estimate.

The Qwen 0.5B record is a small GGUF example. Use a file the app currently offers to download. If it shows no compatible choice, use Ollama's small-model example.

3. Download and let the check finish

Select Download on the model's detail screen. Keep the app open until the download and SHA-256 check finish. The hash check compares the file with the catalog record; it does not judge the model's answers.

If a download stops, open Downloads to check its state and resume when offered. See common fixes if it cannot continue.

4. Open Library and choose Chat

Find the downloaded model in Library and choose Chat. Wait for it to load, then ask: “Explain a rainbow in three short sentences.” You should see a reply appear. If loading fails, try the memory and file-type fixes before downloading a larger model.

5. Try it offline

After the model has loaded and answered once, disconnect Wi-Fi or Ethernet and send another simple message. Reconnect when you want to browse or download again. See what stays on your computer for the limits.