Gemini (SDK / protocol)

Native Gemini protocol (Google GenAI SDK / compatible apps).

Setup steps

1. We expose a native Gemini-compatible endpoint: path /v1beta/models/{model}:generateContent, auth via the x-goog-api-key header or ?key= with your sk-....
2. With the Google GenAI SDK (Python), just point the base URL at us:

from google import genai
from google.genai import types

client = genai.Client(
    api_key="sk-YOUR_KEY",
    http_options=types.HttpOptions(base_url="https://api.llmapi.pro:99"),
)
resp = client.models.generate_content(
    model="gemini-3.1-pro",
    contents="Hello",
)
print(resp.text)

3. Or verify with curl:

curl "https://api.llmapi.pro:99/v1beta/models/gemini-3.1-pro:generateContent" \
  -H "x-goog-api-key: sk-YOUR_KEY" \
  -H "content-type: application/json" \
  -d '{"contents":[{"parts":[{"text":"Hello"}]}]}'

Notes

• The SDK base URL has no /v1beta (the SDK appends /v1beta/models/...); only spell out the full path for manual curl.
• On the Gemini endpoint use a Gemini model name (e.g. gemini-3.1-pro), powered by our top engine; the Gemini protocol is text chat for now (images / tool-calls not yet mapped).
• Troubleshooting: an empty reply with HTTP 200 usually means an invalid key or quota issue — check your key in the Console first.
• ⚠️ Google's official gemini CLI currently has no switch to change its backend. For a coding CLI use Claude Code or Codex; this page is for the Gemini SDK and apps that allow a custom Gemini base URL.