Gemini SDK

Use /v1beta/models on Triw with the Google Gen AI SDK or REST.

Triw exposes the Gemini REST surface under /v1beta, plus an OpenAI-style route for models.

Install

Python

pip install google-genai

Configure

Setting	Official Google	Triw
Host	https://generativelanguage.googleapis.com	https://router.triw.io
Path	/v1beta/models/{model}:generateContent	same
Key	Google API key	sk-... from Triw, via x-goog-api-key or Authorization

curl "https://router.triw.io/v1beta/models/gemini-2.5-flash:generateContent" \
  -H "x-goog-api-key: $TRIW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{"parts": [{"text": "Explain vector search in two sentences."}]}]
  }'

Streaming uses the standard method suffix:

curl "https://router.triw.io/v1beta/models/gemini-2.5-flash:streamGenerateContent?alt=sse" \
  -H "x-goog-api-key: $TRIW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"contents":[{"parts":[{"text":"Count to five."}]}]}'

With the Google Gen AI SDK

from google import genai

client = genai.Client(
    api_key="sk-...",
    http_options={"base_url": "https://router.triw.io"},
)

resp = client.models.generate_content(
    model="gemini-2.5-flash",
    contents="Explain consistent hashing in two sentences.",
)
print(resp.text)

Exact option names differ between SDK generations (google-generativeai vs google-genai); the values are always host https://router.triw.io and your sk- key.

OpenAI-style access to the same models

If your stack is already OpenAI-shaped, skip the Gemini protocol:

curl https://router.triw.io/v1/chat/completions \
  -H "Authorization: Bearer $TRIW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-2.5-flash","messages":[{"role":"user","content":"ping"}]}'

Model discovery for OpenAI-style clients is available at GET /v1/models; the Gemini-native listing lives under /v1beta/openai/models.