extract() Method

// Get page text
const { pageText } = await stagehand.extract();

// Answer a question
const { extraction } = await stagehand.extract("What is the article about?");

// Extract with custom schema
const data = await stagehand.extract(instruction, schema, options?);

await stagehand.extract();
await stagehand.extract(options?);

await stagehand.extract(instruction, options?);

await stagehand.extract(instruction, schema, options?);

import { Stagehand } from "@browserbasehq/stagehand";

const stagehand = new Stagehand({ env: "LOCAL" });
await stagehand.init();

const page = await stagehand.context.newPage();
await page.goto("https://example.com");

// Get all text content
const { pageText } = await stagehand.extract();
console.log(pageText);

await stagehand.close();

await page.goto("https://news.ycombinator.com");

// Extract specific information
const { extraction } = await stagehand.extract(
  "What is the title of the top story?"
);

console.log(extraction); // "New AI Framework Released"

import { z } from "zod";

await page.goto("https://example-shop.com/product/123");

// Define schema
const productSchema = z.object({
  name: z.string(),
  price: z.string(),
  description: z.string(),
  inStock: z.boolean(),
  rating: z.number().optional(),
});

// Extract data
const product = await stagehand.extract(
  "Extract the product details",
  productSchema
);

console.log(product);
// {
//   name: "Wireless Mouse",
//   price: "$29.99",
//   description: "Ergonomic wireless mouse with...",
//   inStock: true,
//   rating: 4.5
// }

import { z } from "zod";

await page.goto("https://example.com/articles");

const articleListSchema = z.object({
  articles: z.array(
    z.object({
      title: z.string(),
      author: z.string(),
      date: z.string(),
      summary: z.string().optional(),
    })
  ),
});

const { articles } = await stagehand.extract(
  "Extract all articles from the page",
  articleListSchema
);

console.log(`Found ${articles.length} articles`);

// Extract only from specific section
const headerData = await stagehand.extract(
  "Get the navigation links",
  z.object({
    links: z.array(z.object({ text: z.string(), url: z.string() })),
  }),
  { selector: "header nav" }
);

import { z } from "zod";

const jobSchema = z.object({
  title: z.string().describe("Job title"),
  company: z.string().describe("Company name"),
  location: z.string().describe("Job location"),
  salary: z
    .string()
    .optional()
    .describe("Salary range if available"),
  remote: z.boolean().describe("Whether the job is remote"),
  requirements: z
    .array(z.string())
    .describe("List of job requirements"),
});

await page.goto("https://jobs.example.com/posting/123");

const job = await stagehand.extract(
  "Extract the job posting details",
  jobSchema
);

// Use different model for extraction
const data = await stagehand.extract(
  "Extract contact information",
  contactSchema,
  {
    model: "anthropic/claude-3-5-sonnet-latest",
  }
);

const page1 = await stagehand.context.newPage();
const page2 = await stagehand.context.newPage();

await page1.goto("https://example.com/page1");
await page2.goto("https://example.com/page2");

// Extract from specific pages
const data1 = await stagehand.extract("Get the title", schema, { page: page1 });
const data2 = await stagehand.extract("Get the title", schema, { page: page2 });

const productSchema = z.object({
  product: z.object({
    name: z.string(),
    brand: z.string(),
    price: z.object({
      current: z.string(),
      original: z.string().optional(),
      currency: z.string(),
    }),
    availability: z.enum(["in_stock", "out_of_stock", "pre_order"]),
    images: z.array(z.string().url()),
    specifications: z.record(z.string(), z.string()),
    reviews: z.object({
      averageRating: z.number(),
      totalReviews: z.number(),
    }).optional(),
  }),
});

const data = await stagehand.extract(
  "Extract complete product information",
  productSchema
);

const newsSchema = z.object({
  article: z.object({
    headline: z.string(),
    subheading: z.string().optional(),
    author: z.string(),
    publishDate: z.string(),
    content: z.string(),
    tags: z.array(z.string()),
    relatedArticles: z.array(
      z.object({
        title: z.string(),
        url: z.string(),
      })
    ).optional(),
  }),
});

const article = await stagehand.extract(
  "Extract the article content and metadata",
  newsSchema
);

const contactSchema = z.object({
  contact: z.object({
    email: z.string().email().optional(),
    phone: z.string().optional(),
    address: z.object({
      street: z.string(),
      city: z.string(),
      state: z.string(),
      zip: z.string(),
      country: z.string(),
    }).optional(),
    socialMedia: z.object({
      twitter: z.string().optional(),
      linkedin: z.string().optional(),
      facebook: z.string().optional(),
    }).optional(),
  }),
});

const contact = await stagehand.extract(
  "Extract all contact information",
  contactSchema
);