Google Gemini cost estimator (node)

estimator.js

const {GoogleGenerativeAI} = require('@google/generative-ai');
/**
 * Estimates the cost of a Gemini API invocation based on token usage.
 *
 * @param {Object} usageMetadata - The usage metadata from the API response.
 * @param {number} usageMetadata.promptTokenCount - The number of tokens in the input prompt.
 * @param {number} usageMetadata.candidatesTokenCount - The number of tokens in the output.
 * @returns {number} The estimated cost of the invocation in USD.
 *
 * Pricing information:
 * - For prompts up to 128k tokens:
 *   - Input: $0.075 per 1 million tokens
 *   - Output: $0.30 per 1 million tokens
 * - For prompts longer than 128k tokens:
 *   - Input: $0.15 per 1 million tokens
 *   - Output: $0.60 per 1 million tokens
 *
 * Source: [Google AI Pricing](https://ai.google.dev/pricing#1_5flash)
 */
function estimateInvocationCost(usageMetadata) {
  const { promptTokenCount, candidatesTokenCount } = usageMetadata;
  const totalTokenCount = promptTokenCount + candidatesTokenCount;

  let inputCostPerMillionTokens;
  let outputCostPerMillionTokens;

  if (totalTokenCount <= 128000) {
      inputCostPerMillionTokens = 0.075;
      outputCostPerMillionTokens = 0.30;
  } else {
      inputCostPerMillionTokens = 0.15;
      outputCostPerMillionTokens = 0.60;
  }

  const inputCost = (promptTokenCount / 1e6) * inputCostPerMillionTokens;
  const outputCost = (candidatesTokenCount / 1e6) * outputCostPerMillionTokens;

  return inputCost + outputCost;
}

const model = client.getGenerativeModel({
    model: 'gemini-1.5-flash', 
    safetySettings: safetySettings,
    generation_config:{
        "response_mime_type": "application/json",
        "temperature": 0.0
    }
});
const result = await model.generateContent(prompt);
const usageMetadata = result.response.usageMetadata;

const cost = estimateInvocationCost(usageMetadata);
console.log(`Estimated cost of parsing stripped ${url}: $${cost.toFixed(6)}`);