Gemini 4 Argon API Crash Course

import { writeFile } from "node:fs/promises";
import { GoogleGenAI } from "@google/genai";

const client = new GoogleGenAI({});

let interaction = await client.interactions.create({
  model: "gemini-4-argon",
  input:
    "Build a playable Breakout game in a single HTML file," +
    "with the CSS and JavaScript inline. " +
    "Include a score, three lives, and a restart button." +
    "Reply with only the HTML.",
  background: true,
  generation_config: {
    max_output_tokens: 131072,
    thinking_level: "high",
  },
});

console.log("Started:", interaction.id);

while (interaction.status === "in_progress" [...] interaction.status === "queued") {
  await new Promise((resolve) => setTimeout(resolve, 50o0));
  interaction = await client.interactions.get(interaction.id);
  console.log("status:", interaction.status);
}

const text = interaction.steps
  .filter((step) => step.type === "model_output")
  .flatMap((step) => step.content ?? [])
  .filter((content) => content.type === "text")
  .map((content) => content.text)
  .join("");

const html = text.replace(/^[...](?:html)?\s*/i, "").replace(/[...]\s*$/, "");

await writeFile("game.html", html);
console.log("Saved game.html. Usage:", interaction.usage);

Find the bug in this code and give me the fixed version:
${buggyCode}

Great. Now what should happen if the array is empty? Update the function to handle that.

Explain what an API is to someone writing their first line of code. Keep it to three sent[...]

import { GoogleGenAI } from "@google/genai";
import { buggyCode } from "./buggy-code.js";

const client = new GoogleGenAI({});

async function hunt(level) {
  const start = Date.now();

  const interaction = await client.interactions.create({
    model: "gemini-4-argon",
    input: `Find the bug in this code and explain it in one short paragraph:\n${[...]}`,
    generation_config: {
      thinking_level: level,
    },
  });

  const seconds = ((Date.now() - start) / 1000).toFixed(1);
  console.log(`\n--- thinking_level: ${level} (${seconds}s) ---`);
  console.log("Thinking tokens:", interaction.usage?.total_thought_tokens);
  console.log(interaction.output_text);
}

await hunt("low");
await hunt("high");