Skip to main content

curl quickstart

Check the full request path with a minimal OpenAI compatible call.

Before you begin​

  • Prepare YOUR_API_KEY on the keys page.
  • Choose an available model on the models and pricing page and substitute it for YOUR_MODEL.
  • Confirm that your network can reach https://api.popgo.site.

View live models and pricing

Preflight checklist​

  • Replace only YOUR_API_KEY and YOUR_MODEL first. Do not change path, protocol, and advanced client options at the same time.
  • If the app asks for a Base URL, enter the root or /v1 level from its guide. Only enter a full Endpoint when the app explicitly asks for one.
  • Get one successful quickstart response, then copy the same Key, model, and address into the client.

Choose the address and endpoint by protocol​

OpenAI compatible clients usually use the /v1 Base URL. Claude Messages and native Gemini use the root address, then the client or full Endpoint adds the protocol path.

See API Reference for the complete public endpoint list. See Image and Video Generation for video tasks and media parameters.

Local environment preparation​

If you will run a CLI, automation script, or tool that needs Node.js, install Node.js and npm first. Check versions in the same terminal that starts the tool:

node --version
npm --version
  • Windows: use the official installer or a system package manager for the Node.js LTS release.
  • macOS: use the official installer or Homebrew.
  • Linux/WSL2: use the distribution package manager or the Node.js recommended installation method.
  • If the command is still missing after installation, close the old terminal and open a new one so the updated PATH is loaded.

Send the request​

This request uses the OpenAI compatible Base URL.

curl https://api.popgo.site/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_MODEL",
    "messages": [
      {"role": "user", "content": "Reply with a connection confirmation"}
    ]
  }'

What success looks like​

A successful HTTP status with assistant content confirms the Base URL, key, and model. If it fails, continue with the matching status code in troubleshooting.

Next step​

Continue with the OpenAI, Claude, or Gemini protocol guide supported by your client, or choose a software integration.