Google Gemini

Gemini is the fastest Google model to put behind Sentisec: one API key from Google AI Studio, no cloud project, no service account. Sentisec speaks Gemini's OpenAI-compatible chat-completions surface, so any OpenAI-shaped client works with a base-URL change and nothing else.

If your deployment needs to live inside a Google Cloud project — your own IAM, region, and billing — use Vertex AI instead.

Setup

Get an API key from aistudio.google.com. It looks like AIza….

Add it to your workspace. In the dashboard, open Settings → Providers. Under Provider keys, choose Add Gemini key on the Google Gemini API card and paste the key exactly as Google gave it to you — no wrapper, no JSON.

Mint a Sentisec API key. Settings → API keys → Issue key. Format: sk_sentisec_<8>_<32>.

Point your client at your workspace endpoint.

export OPENAI_BASE_URL=https://api.sentisec.ch/ws/<your-workspace-id>/v1
export OPENAI_API_KEY=$SENTISEC_API_KEY

The credential takes effect on your next request — nothing to restart.

First call

Any model id starting with gemini- routes to Gemini automatically. No header needed.

curl -X POST https://api.sentisec.ch/ws/<your-workspace-id>/v1/chat/completions \
  -H "Authorization: Bearer $SENTISEC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.6-flash",
    "messages": [{"role": "user", "content": "hi"}]
  }'

The response is the standard OpenAI Chat Completions body, and the session appears under Sessions in the dashboard.

OpenAI SDKs

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.sentisec.ch/ws/<your-workspace-id>/v1",
    api_key=os.environ["SENTISEC_API_KEY"],
)

resp = client.chat.completions.create(
    model="gemini-3.6-flash",
    messages=[{"role": "user", "content": "hi"}],
)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.sentisec.ch/ws/<your-workspace-id>/v1",
  apiKey: process.env.SENTISEC_API_KEY!,
});

const resp = await client.chat.completions.create({
  model: "gemini-3.6-flash",
  messages: [{ role: "user", content: "hi" }],
});

Streaming ("stream": true) and OpenAI-format tool calling both work, and tool calls are gated the same way they are on every other provider.

Model ids

Model id	Notes
`gemini-3.6-flash`	Fast general-purpose; the usual default.
`gemini-2.5-flash-image`	Image-capable.
`gemini-3-pro-image-preview`	Preview image model.
`gemini-embedding-001`	Embeddings — see below.

Any id starting with gemini- is forwarded to Google, which is the source of truth for what is available to your key. The table above is what appears in model discovery (GET .../v1/models); ids outside it still dispatch.

Embeddings

Gemini is one of the two providers that serve POST .../v1/embeddings through Sentisec:

curl -X POST https://api.sentisec.ch/ws/<your-workspace-id>/v1/embeddings \
  -H "Authorization: Bearer $SENTISEC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "gemini-embedding-001", "input": "hello"}'

Embeddings relay verbatim and are recorded for usage; there is no tool call in an embedding to gate.

If something goes wrong

Response	What it means
`401` / `402 no_upstream_credential`	No Gemini key is stored on this workspace. Add it under Settings → Providers, in Provider keys — the request's row under Requests offers Add Gemini key, which opens that page with Gemini selected.
A Google `400 API_KEY_INVALID`	The key was revoked or mistyped at AI Studio. Re-paste it.
Request went to a different provider	Your model id does not start with `gemini-`. Either use a `gemini-` id or send `X-Sentisec-Upstream-Provider: gemini`.

More: Provider unavailable.