n8n Workflow: AI Agent Token Usage and Cost Estimator

This n8n workflow provides a robust solution for tracking token usage and estimating costs for various AI agents (OpenAI, Anthropic, Google Gemini) and logging this data to Google Sheets. It's designed to be invoked as a sub-workflow, making it a reusable component for any workflow that interacts with AI models.

What it does

This workflow acts as a centralized logging and cost estimation service for AI agent interactions. Specifically, it performs the following steps:

1. Receives Data: It's triggered by another workflow, expecting input containing details about an AI agent's execution, including the model used, input tokens, output tokens, and an estimated cost.
2. Calculates Total Tokens: It sums the input and output tokens to get a total token count for the AI agent's interaction.
3. Estimates Cost (Placeholder): The workflow is set up to receive an estimated cost, implying that the calling workflow is responsible for calculating this based on model pricing.
4. Logs to Google Sheets: It appends a new row to a specified Google Sheet, recording the AI agent's name, model, total tokens, and estimated cost.
5. Returns Data: It passes the processed data back to the calling workflow.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance to import and execute the workflow.
• Google Sheets Account: A Google account with access to Google Sheets.
• Google Sheets Credential: An n8n credential configured for Google Sheets (OAuth2 recommended) with access to the target spreadsheet.
• AI Agent Workflow: Another n8n workflow that uses AI agents (OpenAI, Anthropic, Google Gemini) and can pass relevant token usage and cost data to this sub-workflow.

Setup/Usage

1. Import the Workflow:
   • Download the provided JSON file for this workflow.
   • In your n8n instance, go to "Workflows" and click "New".
   • Click the "Import from JSON" button and paste the workflow JSON or upload the file.
2. Configure Google Sheets:
   • Locate the "Google Sheets" node in the workflow.
   • Select or create a Google Sheets credential.
   • Specify the Spreadsheet ID and Sheet Name where you want to log the AI agent data.
3. Integrate as a Sub-workflow:
   • In your main workflow (the one using AI agents), add an "Execute Workflow" node.
   • Configure the "Execute Workflow" node to call this "AI Agent Token Usage and Cost Estimator" workflow.
   • Ensure your main workflow passes the following data to this sub-workflow:
     • agentName (string): The name of the AI agent.
     • model (string): The AI model used (e.g., "gpt-4", "claude-3-opus", "gemini-pro").
     • inputTokens (number): The number of tokens used for the input prompt.
     • outputTokens (number): The number of tokens generated in the response.
     • estimatedCost (number): The calculated cost for the AI interaction.
   • The "Edit Fields" node in this sub-workflow expects these fields to be present in the input.

Example of input expected by this sub-workflow:

{
  "agentName": "MyChatbot",
  "model": "gpt-4",
  "inputTokens": 100,
  "outputTokens": 50,
  "estimatedCost": 0.003
}

This workflow will then process this information and append it to your designated Google Sheet.