AI & Machine Learning

Google Gemini Integration

Google Gemini API client for AI text generation, chat conversations, and content creation. Provides access to Google's advanced language models for various generative AI tasks.

Google Gemini

Category: AI & Machine Learning
Provider Key: googleGemini
SDK Packages: @google/generative-ai@^0.24.1

Google Gemini API client for AI text generation, chat conversations, and content creation. Provides access to Google's advanced language models for various generative AI tasks.

Configuration

To use Google Gemini in your project, add it to your project integrations and provide the following configuration:

Parameter	Type	Required	Description
`apiKey`	`string`	Yes	Google AI Studio API key
`model`	`string`	No	Model to use (e.g., 'gemini-pro', 'gemini-pro-vision')
`temperature`	`number`	No	Temperature for generation (0.0 to 1.0)
`maxOutputTokens`	`number`	No	Maximum tokens in response
`topP`	`number`	No	Top-p sampling value
`topK`	`number`	No	Top-k sampling value

Example Configuration

{
  "provider": "googleGemini",
  "configuration": [
    { "name": "apiKey", "value": "your-apiKey" },
    { "name": "model", "value": "your-model" },
    { "name": "temperature", "value": 0 },
    { "name": "maxOutputTokens", "value": 0 },
    { "name": "topP", "value": 0 },
    { "name": "topK", "value": 0 }
  ]
}

Available Methods

Quick reference:

Generation: generateContent, generateContentStream
Chat: startChat
Utilities: countTokens
Embeddings: embedContent, batchEmbedContents

Generation

`generateContent`

Generate Content

Generate text content from a prompt

Parameters:

Parameter	Type	Required	Description
`prompt`	`string`	Yes	The text prompt for generation
`images`	`Array`	No	Array of image data (base64 or URLs) for vision models
`safetySettings`	`Object`	No	Safety settings for content filtering

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "generateContentAction",
  "provider": "googleGemini",
  "action": "generateContent",
  "parameters": [
    { "parameterName": "prompt", "parameterValue": "'your-prompt'" },
    { "parameterName": "images", "parameterValue": "[]" },
    { "parameterName": "safetySettings", "parameterValue": "{}" }
  ],
  "contextPropertyName": "generateContentResult"
}

MScript example:

await _googleGemini.generateContent({
  prompt: /* string */,
  images: /* Array */,
  safetySettings: /* Object */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.generateContent({
  prompt: /* string */,
  images: /* Array */,
  safetySettings: /* Object */,
});

`generateContentStream`

Generate Content Stream

Stream generated content for longer responses

Parameters:

Parameter	Type	Required	Description
`prompt`	`string`	Yes	The text prompt for generation
`onChunk`	`Function`	Yes	Callback function for each streamed chunk

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "generateContentStreamAction",
  "provider": "googleGemini",
  "action": "generateContentStream",
  "parameters": [
    { "parameterName": "prompt", "parameterValue": "'your-prompt'" },
    { "parameterName": "onChunk", "parameterValue": "'your-onChunk'" }
  ],
  "contextPropertyName": "generateContentStreamResult"
}

MScript example:

await _googleGemini.generateContentStream({
  prompt: /* string */,
  onChunk: /* Function */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.generateContentStream({
  prompt: /* string */,
  onChunk: /* Function */,
});

Chat

`startChat`

Start Chat

Start a multi-turn chat conversation

Parameters:

Parameter	Type	Required	Description
`history`	`Array`	No	Previous conversation history
`generationConfig`	`Object`	No	Generation configuration for this chat

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "startChatAction",
  "provider": "googleGemini",
  "action": "startChat",
  "parameters": [
    { "parameterName": "history", "parameterValue": "[]" },
    { "parameterName": "generationConfig", "parameterValue": "{}" }
  ],
  "contextPropertyName": "startChatResult"
}

MScript example:

await _googleGemini.startChat({
  history: /* Array */,
  generationConfig: /* Object */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.startChat({
  history: /* Array */,
  generationConfig: /* Object */,
});

Utilities

`countTokens`

Count Tokens

Count tokens in a text prompt

Parameters:

Parameter	Type	Required	Description
`text`	`string`	Yes	Text to count tokens for

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "countTokensAction",
  "provider": "googleGemini",
  "action": "countTokens",
  "parameters": [
    { "parameterName": "text", "parameterValue": "'your-text'" }
  ],
  "contextPropertyName": "countTokensResult"
}

MScript example:

await _googleGemini.countTokens({
  text: /* string */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.countTokens({
  text: /* string */,
});

Embeddings

`embedContent`

Embed Content

Embed text for semantic search or similarity

Parameters:

Parameter	Type	Required	Description
`content`	`string`	Yes	Text content to embed
`taskType`	`string (`"RETRIEVAL_QUERY"`,` "RETRIEVAL_DOCUMENT"`,` "SEMANTIC_SIMILARITY"`,` "CLASSIFICATION"`,` "CLUSTERING"`)`	No	Task type for embedding (e.g., 'RETRIEVAL_QUERY', 'RETRIEVAL_DOCUMENT')
`model`	`string`	No	Embedding model to use (default: 'embedding-001')

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "embedContentAction",
  "provider": "googleGemini",
  "action": "embedContent",
  "parameters": [
    { "parameterName": "content", "parameterValue": "'your-content'" },
    { "parameterName": "taskType", "parameterValue": "'your-taskType'" },
    { "parameterName": "model", "parameterValue": "'your-model'" }
  ],
  "contextPropertyName": "embedContentResult"
}

MScript example:

await _googleGemini.embedContent({
  content: /* string */,
  taskType: /* string */,
  model: /* string */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.embedContent({
  content: /* string */,
  taskType: /* string */,
  model: /* string */,
});

`batchEmbedContents`

Batch Embed Contents

Batch embed multiple texts

Parameters:

Parameter	Type	Required	Description
`contents`	`Array<string>`	Yes	Array of text contents to embed
`taskType`	`string (`"RETRIEVAL_QUERY"`,` "RETRIEVAL_DOCUMENT"`,` "SEMANTIC_SIMILARITY"`,` "CLASSIFICATION"`,` "CLUSTERING"`)`	No	Task type for embedding
`model`	`string`	No	Embedding model to use

IntegrationAction example:

{
  "extendClassName": "IntegrationAction",
  "name": "batchEmbedContentsAction",
  "provider": "googleGemini",
  "action": "batchEmbedContents",
  "parameters": [
    { "parameterName": "contents", "parameterValue": "'your-contents'" },
    { "parameterName": "taskType", "parameterValue": "'your-taskType'" },
    { "parameterName": "model", "parameterValue": "'your-model'" }
  ],
  "contextPropertyName": "batchEmbedContentsResult"
}

MScript example:

await _googleGemini.batchEmbedContents({
  contents: /* Array<string> */,
  taskType: /* string */,
  model: /* string */,
})

Service library example:

const { getIntegrationClient } = require("integrations");
const client = await getIntegrationClient("googleGemini");
const result = await client.batchEmbedContents({
  contents: /* Array<string> */,
  taskType: /* string */,
  model: /* string */,
});

Was this page helpful?

Last updated today

Built with Documentation.AI