No matches. Try a model format, tool name or error.
On this page
Start using Hugging Bay
Run an open AI model on your own computer. Choose a supported model and an app to run it.
1. Install the app
Download the Hugging Bay app for your computer. The app and model files are separate downloads, so you can choose your model.
2. Choose a model
Go to Browse, then pick your memory and your task, or look through the whole catalog. Check the model’s file size and runtime requirements before downloading.
- Likely fits is a memory estimate for the selected file.
- Tight fit means you may need a smaller file.
- Check fit means more information is needed.
3. Download and chat
Open a model page to see its current download options. Download a supported file, verify its SHA-256, then open it in your app. Your first reply can take longer while the model loads.
# Prefer the terminal? Search the public catalog:
curl 'https://huggingbay.xyz/api/v1/search?q=qwen&limit=5'
# See /developers/cli for the installed CLI's commands.Where your models live
The desktop app keeps its own local files. My library on this website shows saved models and downloads managed in this browser. Opening a download link alone does not confirm the file was saved.
Get started
Find your first step.
- Start here
Choose a way to use Hugging Bay.
- Your first local model
Go from choosing a model to your first reply.
- Your first local model
Install the desktop app or run a model from the terminal.
- Developer quickstart
Connect an assistant or make your first API call.
- Learn the basics
Guides to open models.
Run models
Choose files and check how to use them.
- Model guides
Answers about model formats, memory and downloads.
- Hugging Face downloads
Set HF_ENDPOINT once. Commands for hf, Python and llama.cpp, with live test results.
- Download one Hugging Face file
Pick a hosted file and copy its download command.
Desktop app
Get the app and open your downloads.
- Open your download
Open model files in a local app.
- Get the desktop app
Download the app and check its requirements.
- Desktop app help
Install, chat, sharing and fixes for the desktop app.
For developers
Work with the API, CLI and SDK.
- Developer overview
Find the CLI and MCP distribution details.
- API reference
Browse the OpenAPI contract.
- API recipes
Read API examples and curl commands.
- Command line
Install and use the command-line tools.
- Python SDK
Use the Python client.
- API keys
Choose the key for your workflow.
- Installation
Check the available distribution channels.
- Integration status
Check the status of integration guides.
- Download protocol
Resolve files, download exact bytes and verify checksums.
- Build with Hugging Bay
Guides for building applications.
For agents
Connect an assistant and inspect the catalog.
- Atom and RSS feeds
Follow public catalog changes and dated file events.
- Agent client mirrors
HF_ENDPOINT mirror and Ollama pull routes inside existing clients.
- Connect an assistant
Set up MCP or OpenAPI in your assistant.
- MCP reference
Install MCP and inspect the remote endpoint.
- Agent cookbook
Inspect model records and their evidence.
- Build an agent
Connect tools and build a model workflow.
- Agent guardrails
Verify results before completing a task.
- Agent access
Find the public agent contracts.
- AI search guidance
Find llms.txt, crawl guidance and citation sources.
- Agent recommendation kit
Read integration examples and claim boundaries.
- Agent blackboard
Read the shared agent board contract.
Teach
Prepare to train a model locally.
- Teach locally
Choose a starting model for local training.
Troubleshooting
Get help with files, checks and errors.
- Help
Find steps for common download and model errors.
- File verification API
Look up a file fingerprint in the hosted index.
- Download verification
Read how file checks and evidence work.