DeepSeek Harness
Switch Harness to an OpenAI-compatible provider and use DeepSeek models through XycAi.
One-click setup with our desktop client
Download the XycAi client to install and configure everything automatically — no manual steps needed.
Connection values
Open Harness Provider / Model settings and choose OpenAI Compatible:
Base URL: https://apicdn.xyc.ai/v1
API Key: sk-xxxxxxxx
Model: deepseek-v4-proVerify the connection
Save the provider, run a short task, and confirm a successful request in the logs.
Troubleshooting
- Keep the base URL at /v1; Harness appends the request path.
- If the model list is empty, enter the model name manually.