Google Gemini

Google's Gemini API: generate text, list models, count tokens and create embeddings.

5 actions API key 5 read0 write Calls generativelanguage.googleapis.com

Connect

TypeScript

const conn = await setu.connections.initiate({ userId: "user_1", toolkit: "gemini" });
console.log(conn.connectUrl); // show this to your user
await conn.waitUntilActive();

Permissions requested

This toolkit uses a single credential and requests no scopes.

Actions

Count tokensreadgemini.count_tokensCount how many tokens a text prompt uses for a given Gemini model.

Input	Type		Description
`model`	string	Required	Model id without the models/ prefix
`text`	string	Required	The text to count

TypeScript

await setu.execute({
  userId: "user_1",
  action: "gemini.count_tokens",
  input: {
    "model": "<model>",
    "text": "<text>"
  },
});

Create embeddingreadgemini.embed_contentCreate a text embedding vector with a Gemini embedding model.

Input	Type		Description
`model`	string	Required	Embedding model id without the models/ prefix, e.g. gemini-embedding-001
`text`	string	Required	Text to embed
`taskType`	string	Optional	Optional task type, e.g. RETRIEVAL_QUERY
`outputDimensionality`	integer	Optional	Optional vector size

TypeScript

await setu.execute({
  userId: "user_1",
  action: "gemini.embed_content",
  input: {
    "model": "<model>",
    "text": "<text>"
  },
});

Generate textreadgemini.generate_contentGenerate a text response from a Gemini model for a prompt, with an optional system instruction.

Input	Type		Description
`model`	string	Required	Model id without the models/ prefix, e.g. gemini-2.5-flash
`prompt`	string	Required	The user prompt
`systemInstruction`	object	Optional	Optional system instruction as {parts: [{text: ...}]}
`generationConfig`	object	Optional	Optional settings such as temperature and maxOutputTokens

TypeScript

await setu.execute({
  userId: "user_1",
  action: "gemini.generate_content",
  input: {
    "model": "<model>",
    "prompt": "<prompt>"
  },
});

Get modelreadgemini.get_modelGet one Gemini model's details: display name, token limits and supported generation methods.

Input	Type		Description
`model`	string	Required	Model id without the models/ prefix

TypeScript

await setu.execute({
  userId: "user_1",
  action: "gemini.get_model",
  input: {
    "model": "<model>"
  },
});

List modelsreadgemini.list_modelsList the Gemini models available to this key with their token limits and supported methods.

Input	Type		Description
`pageSize`	integer	Optional	Models per page (default 50, max 1000)
`pageToken`	string	Optional	Token from a previous response

TypeScript

await setu.execute({
  userId: "user_1",
  action: "gemini.list_models",
  input: {},
});