Docs / Strand / connectors/ai-google

Google Gemini Connector

Send prompts to Google Gemini models and receive completions from your Strand workflows. Use this connector for summarization, classification, content generation, data extraction, and any task that benefits from a large language model.

Direction: Read / Write | Type: ai.google

Prerequisites

You need a Google AI API key:

1. Go to aistudio.google.com/apikey
2. Click Create API key
3. Select or create a Google Cloud project
4. Copy the generated key

This key uses the Google Generative Language API (Google AI Studio), not Vertex AI. No Google Cloud billing setup is required for the free tier.

Connector Setup

Create a Google Gemini connector from the Connectors page.

Configuration Fields

Operations

Chat Completion

Generate a chat completion response from a Gemini model. This is the single operation exposed by the connector.

Operation ID: chat

The prompt is sent as a user message. If System Instructions are configured on the node, they are sent as the system instruction preceding the prompt.

Models

You can use any valid Google AI model name.

Node Configuration

When you drag a Google Gemini connector node onto the canvas, these fields are available in the node inspector. They apply to this specific node and override connector-level defaults where noted.

Output

A successful chat completion returns:

{
  "response": "Here is a summary of the provided data...",
  "model": "gemini-2.0-flash",
  "provider": "google",
  "usage": {
    "prompt_tokens": 150,
    "completion_tokens": 89,
    "total_tokens": 239
  }
}

When memory is enabled, the output includes "memory_enabled": true. When MCP tools are called, the output includes an "mcp_tools_called" array. If retries occurred, a "retries" count is included.

Access the response in downstream nodes:

{{ steps.NODE_ID.output_payload.response }}
{{ steps.NODE_ID.output_payload.usage.total_tokens }}

Example Workflow

Extract structured data from incoming emails:

1. HTTP Trigger receives a webhook with { "subject": "...", "body": "...", "from": "..." }
2. Google Gemini node extracts fields:

• System Instructions: Extract structured data from the email. Return valid JSON with keys: sender_name, company, intent (one of: inquiry, complaint, order, other), urgency (low, medium, high), and summary (one sentence).
• Prompt: From: {{ payload.from }}\nSubject: {{ payload.subject }}\n\n{{ payload.body }}
• Temperature: 0
• Model: gemini-2.0-flash

1. Python Snippet node parses the JSON response
2. Router node branches on the extracted intent and urgency fields to route the email

Retries

The node automatically retries on transient errors -- rate limits (429), server errors (5xx), timeouts, and connection failures. Retries use exponential backoff with jitter and respect Retry-After headers from Google. Non-retryable errors (invalid API key, bad request) fail immediately.

Conversation Memory

Toggle Memory Enabled on and the node will remember every exchange across workflow runs. Each node gets its own isolated, account-scoped memory that is securely persisted across runs. See AI Connectors for full details on memory behavior, TTL, trimming, and when to enable it.

MCP Tool Access

AI nodes can call external tools via MCP servers. Add one or more MCP server connectors to the MCP Servers field and the model will have access to those tools during its response.

Errors

Related

• AI Connectors -- shared node config, memory, MCP, and examples for all providers
• OpenAI | Anthropic Claude
• MCP Server -- for tool access

← Back to Strand docs