Google Gemini
Call the Gemini Developer API — generate content, count tokens, embed content, list models.
About
Google Gemini ships in the w6w first-party pack. It declares 6 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
- App id
io.w6w.gemini- Version
- 0.2.1
- Author
- w6w
- Licence
- MIT
- Categories
- AI & Machine Learning
Overview
Google’s Gemini Developer API generates and embeds content from Gemini’s models, and this app calls it directly: generate content from a prompt (including multi-turn, multimodal input), embed a single piece of text or a batch of them, count tokens before a call, and list or inspect the available models.
Content is passed through in Gemini’s own request shape rather than flattened to a single string, so conversation history and multimodal input survive intact. This makes the app useful both for one-off generation steps in a workflow and for building a multi-turn exchange across several steps.
This is the Gemini Developer API specifically — the surface Google ships for individual developers — not Vertex AI, Google Cloud’s separate enterprise product for calling the same models. Streaming responses, file uploads, cached content and function-calling tool definitions are not exposed here; a caller who needs the raw model response for a prompt, an embedding, or a token count has everything this app offers.
Build with Google Gemini
Three routes to the same 6 actions. The Workflow tab is generated from Google Gemini's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
Batch Embed Content
batch-embed-contents Generate embedding vectors for several texts in one call, using the same model.
Generate Content
generate-content Generate a model response from text, chat history, or multimodal content.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Google Gemini's real ids.
{
"manifestVersion": "2",
"name": "gemini-example",
"steps": [
{
"id": "batch-embed-contents",
"uses": {
"app": "io.w6w.gemini",
"action": "batch-embed-contents",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"model": "<model>",
"texts": "<texts>"
}
}
]
}Here are some of the things you can do
-
Batch Embed Content
performbatch-embed-contents -
Count Tokens
readcount-tokens -
Embed Content
performembed-content -
Generate Content
performgenerate-content -
Get Model
readget-model
+1 more action available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Google Gemini, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Google Gemini. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "batch-embed-contents",
payload: {
model: "<value>",
texts: "<value>",
// taskType: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value);npm install -g @w6w/cliw6w run conn_YOUR_CONNECTION_ID --action batch-embed-contents --payload '{"model":"<value>","texts":"<value>"}'Give an AI agent Google Gemini — without giving it Google Gemini's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.gemini#batch-embed-contents",
"input": {
"model": "<model>",
"texts": "<texts>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Google Gemini connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
What the agent gets
Credentials it can't read
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
A tool surface scoped to the caller
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
A durable workflow in one call
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Health-aware discovery
Google Gemini's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Google Gemini itself is MIT, and the runtime that executes it is source-available (FSL).
Health checks
Google Gemini declares its own checks, so its health is a property of the app rather than something the host guesses at.