Supported model types
The current version supports two main types of models:
- language macromodel: DeepSeek full series, Llama2 (7B/13B), ChatGLM, etc.
- Multimodal AI: A lightweight visual language model with partial support for image understanding
Model switching operation guide
- Click on the "Model Selection" drop-down menu in the upper left corner of the screen.
- Select the target model from the list of installed models
- The system automatically loads the model weights (may take 1-2 minutes to initialize for first time use)
- You can start using it once the status bar shows the "Model Ready" message.
caveat
- Running multiple large models at the same time can take up a lot of memory, so it is recommended to use them separately
- Dialog history is saved independently for different models
- Some features (e.g. network search) may not be available due to model differences
Extended Recommendations
Users can get more third-party adapted models through the "Community Models" section, but should be careful to verify the security of the models. The list of new models will be updated quarterly.
This answer comes from the articleAingDesk: One-Click Installation of AI Models and Chat Interfaces for PCs (with DeepSeek)The































