Skip to content
← Back to guides

DeepSeek Harness setup

DeepSeek Harness lets you add a custom model API in its web interface, using OpenAI or Anthropic protocols.

Before you start

Setup steps

Install from each product’s official download or release page. Choose the Windows package on Windows and the correct Apple Silicon or Intel package on macOS. Existing users can proceed directly to configuration. Labels may vary by version.

  1. Prepare the Node.js environment specified by the official project. Run npx @deepseek-ai/dsh web in your working directory and open the local URL printed in the terminal. Use PowerShell on Windows or a terminal on macOS; official Desktop packages are another option where provided.
  2. Open Settings → Models → Custom model API and add provider ID p6090ti.
  3. Choose OpenAI Completions, OpenAI Responses, or Anthropic Messages. Use the /v1 Base URL for OpenAI formats or the root for Anthropic.
  4. Enter your key, add at least one actual model ID, save, and create a session using that model.
  5. Test text, then any required agent tools.

Configure one protocol per provider and preserve settings before preview-version upgrades. Official model configuration

Troubleshooting

For 401, 404, 429 and other errors, see API reference: Troubleshooting. If that does not solve it, contact us as described in the support and issue-reporting guide. Never send full keys, passwords or verification codes.
QQ user group: 1109931264