The biggest differentiating capability of Libra.app compared to other AI Agent products is localization. The specific features and dependencies are as follows:
- Local Mode: All chats are sent to the local model, requiring the download of a low-bit LLM model optimized for macOS by GreenBitAI, approximately
2.5GB. - Enhanced Mode: Capable of autonomously performing complex tasks such as file searching, web browsing, programming, charting, and report generation. To better protect user's local data and environment, these operations run in an isolated container environment, requiring the download of a container runtime environment, approximately
1GBin size.
If you are in one of the following situations, you may need to configure a proxy to ensure that the local model, Enhanced mode, and local upload of documents in formats such as pdf, docx, xlsx can work properly:
- Corporate intranet
- Mainland China
Configure your proxy software to use global proxy mode, or add the following network addresses to the proxy whitelist rules:
- huggingface.co
- ghcr.io
- docker.io
After configuring the proxy, restart Libra.app and wait about 10 minutes (readiness time depends on your network). You can confirm whether the above issues exist through the Libra.app interface, or verify more specifically by executing the following commands in the macOS Terminal app.
After initialization is complete, it is recommended to disable the proxy.
Note: Some proxy software has features like "virtual network card" or "TUN" mode. These modes need to be disabled when using Libra.
- Confirm if the local model has been downloaded:
du -hd0 ~/.cache/huggingface/hub/models--GreenBitAI--Qwen3-4B-Instruct-2507-layer-mix-bpw-4.0-mlx
If you see the following content, it indicates the local model has been downloaded correctly:
2.5G /Users/libra/.cache/huggingface/hub/models--GreenBitAI--Qwen3-4B-Instruct-2507-layer-mix-bpw-4.0-mlx
- Confirm if the container runtime environment is ready:
/Applications/Libra.app/Contents/Resources/bin/limactl shell libra nerdctl images
If you see the following content, it indicates proper initialization:
REPOSITORY TAG IMAGE ID CREATED PLATFORM SIZE BLOB SIZE
ghcr.io/greenbitai/libra-runner v0.6.3 59727661d104 2 days ago linux/arm64 1.781GB 565.6MB
mcp/markitdown latest ac2cdd96f844 10 days ago linux/arm64 982.2MB 353.9MB
Error message: Loading Local Model
Refer to FAQ
Error message: Execution Engine is not fully ready
Refer to FAQ
Error message: The file content is either empty
Refer to FAQ