DataEyesAI is compatible with the OpenAI, Anthropic, and Gemini API protocols. Any tool that lets you set a custom API address can connect to it.General settings#
First check which protocol the tool uses, then fill in the settings below:| Protocol supported by the tool | Base URL | API key | Typical tools |
|---|
| OpenAI-compatible | https://studio.dataeyes.ai/v1 | sk-xxx | Cursor, Chatbox, Cherry Studio, LobeChat, Dify, various VS Code extensions |
| Anthropic-compatible | https://studio.dataeyes.ai | sk-xxx | Clients that support a custom Anthropic address |
| Gemini-compatible | https://studio.dataeyes.ai | sk-xxx | Clients that support a custom Gemini address |
Enter only the raw key starting with sk-, without Bearer .
Example: Cursor#
Cursor supports custom models only on Pro and higher plans. The free plan only supports Auto mode.
1.
Open Cursor and go to Settings → Models.
2.
Enter sk-xxx as the OpenAI API Key, turn on Override OpenAI Base URL, and enter https://studio.dataeyes.ai/v1.
3.
Click Add Custom Model, enter a model name (for example deepseek-v4.1-flash), and enable it.
4.
Select the model in the chat to start using it.
Troubleshooting#
| Symptom | What to do |
|---|
| 404 / Invalid URL | /v1 is missing or extra. If it fails with /v1, remove it and retry, and vice versa |
| 401 / authentication failed | Check that the key is complete, has no Bearer prefix, and belongs to this site |
| Model not found | The model name must exactly match the Model Hub, and the current key's group must support the model |
| 502 / 504 / timeout | Reasoning models can take a long time; increase the tool's timeout. If it persists, retry later |
If the tool you want to connect isn't listed, contact us and we'll add a guide.