GPT Series: CLI and VS Code / Cursor Deployment Guide

GPT uses the OpenAI-compatible interface. For new projects, verify the Responses API first while keeping Chat Completions as a compatibility option.

1. Use cases

• Use GPT models from CLI tools or custom scripts.
• Configure Continue, Cline, Roo Code, or similar extensions in VS Code / Cursor.
• Any client that supports an OpenAI-compatible Base URL.

2. Core configuration

3. Recommended models

4. CLI environment variables

export OPENAI_API_KEY="your_synterolink_key"
export OPENAI_BASE_URL="https://api.synterolink.com/v1"

If the client only has a Base URL field, enter:

https://api.synterolink.com/v1

5. Verify model list

curl https://api.synterolink.com/v1/models \
  -H "Authorization: Bearer $OPENAI_API_KEY"

Success means the response includes a model list and your target model ID.

6. Verify Responses API

curl https://api.synterolink.com/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
    "model": "gpt-5.4",
    "input": "Reply exactly: OK"
  }'

Note: OpenAI recommends Responses API for new projects; Chat Completions remains useful for compatibility.

7. Verify Chat Completions

curl https://api.synterolink.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
    "model": "gpt-5.4",
    "messages": [
      {"role": "user", "content": "Reply exactly: OK"}
    ]
  }'

8. VS Code + Continue example

models:
  - title: SynteroLink GPT 5.4
    provider: openai
    model: gpt-5.4
    apiBase: https://api.synterolink.com/v1
    apiKey: ${{ env.OPENAI_API_KEY }}

After configuration, ask Continue: Reply exactly: OK.

9. Cursor / Cline / Roo Code configuration

10. Success checklist

• /v1/models returns a model list.
• /v1/responses returns normal text.
• The same model answers inside the editor.
• When it fails, check HTTP status, model ID, and Base URL first.

11. Common errors