Skip to content

AI Provider Configuration ​

Supported Providers ​

Open Note supports 13+ AI providers, including online APIs and local deployments.

Online API Providers ​

  • Best value for money
  • API endpoint: https://api.deepseek.com
  • Model: deepseek-chat

Configuration steps:

  1. Register a DeepSeek account
  2. Get an API Key
  3. Configure it in Open Note settings

OpenAI ​

  • Models: GPT-4, GPT-3.5-turbo
  • API endpoint: https://api.openai.com

Claude (Anthropic) ​

  • Models: Claude 3 Opus/Sonnet/Haiku
  • API endpoint: https://api.anthropic.com

Alibaba Bailian ​

  • Models: Qwen series
  • API endpoint: https://dashscope.aliyuncs.com

Zhipu AI ​

  • Models: GLM-4, GLM-3-Turbo
  • API endpoint: https://open.bigmodel.cn

Other Providers ​

  • Baidu Wenxin
  • Tencent Hunyuan
  • Kimi (Moonshot AI)
  • Volcano Engine

Local Deployment ​

Ollama ​

Fully offline, with data kept local:

bash
# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Download a model
ollama pull llama3

# Configure Open Note
Endpoint: http://localhost:11434
Model: llama3

Configuration Steps ​

  1. Open Settings → AI Services
  2. Select a provider
  3. Enter the API Key (required for online APIs)
  4. Select a model
  5. Test the connection

Model Recommendations ​

Use CaseRecommended ModelNotes
Daily conversationDeepSeek / GPT-3.5Fast and low cost
Complex reasoningGPT-4 / Claude 3Higher quality
Local privacyOllama / local deploymentData stays local
Chinese optimizationQwen / GLMBetter Chinese understanding

Embedding Service ​

Open Note uses a local Embedding service for vectorization:

Automatic Installation ​

Installed automatically on first launch (model files must be downloaded manually).

Manual Model Download ​

Go to Settings - Knowledge Base - Embedding Model, select a model version, and click Download.

Model Source ​

ModelScope Community (modelscope.cn)

Thanks to the ModelScope Community for its contributions to open source and open models.

Troubleshooting ​

Q: Connection timeout ​

Check your network and proxy settings.

Q: Invalid API Key ​

Confirm that the key is correct and has available balance.

Q: Local Embedding failed to start ​

Ensure Python 3.10+ is installed and enough memory is available (2GB+ free memory recommended).