Integrate with Cherry Studio

Cherry Studio integrates with GPUStack to leverage locally hosted LLMs, embeddings, and reranking capabilities.

Deploying Models

1. In GPUStack UI, navigate to the Deployments page and click on Deploy Model to deploy the models you need. Here are some example models:

   • qwen3-instruct-2507
   • qwen2.5-vl-7b
   • bge-m3
   • bge-reranker-v2-m3

2. In the model’s Operations, open API Access Info to see how to integrate with this model:

Create an API Key

1. Navigate to the Access Control > API Keys page, then click on New API Key.

2. Fill in the name, then click Save.

3. Copy the API key and save it for later use.

Integrating GPUStack into Cherry Studio

1. Open Cherry Studio, go to Settings → Model Provider, find GPUStack, enable it, and configure it as shown:

   • API Key: Input the API key you copied from previous steps.

   • API Host: Access URL in the API Access Info panel.

2. In the GPUStack provider configuration, click "Manage" and enable the models you need:

3. (Optional) Test the API:

After configuration, return to the Cherry Studio home page and start using your models.

Using LLMs

Using Multimodal Models

1. Select a multimodal model:

2. Ask multimodal questions:

Use Embeddings and Reranking to Improve Knowledge Base Q&A

1. Open the Knowledge Base configuration page:

2. Add a knowledge base:

3. Add content to the knowledge base (using “Notes” as an example):

4. Return to the home page and use knowledge base Q&A: