Introducing SerpApi Bing Copilot API

Bing Copilot (also known as Copilot Search) is Microsoft’s new AI-powered search assistant. It combines organic and generative search to provide concise, conversational answers to queries. In practice, Bing Copilot returns a summary card with a short answer (often with an illustrative image/video) and a numbered list of source links. SerpApi’s new Bing Copilot API lets you scrape these AI-generated results via a simple endpoint (engine=bing_copilot), returning the content and citations in structured JSON (no need to deal with Bing’s complex page layout or anti-bot measures yourself).

Bing Copilot’s AI answers contain rich, structured data – but scraping them at scale is challenging. The Copilot interface uses dynamic, nested HTML and frequent updates, plus anti-bot protections. SerpApi’s Bing Copilot API solves this by handling the heavy lifting (proxies, captchas, layout changes) and giving you reliable JSON output.

What is Bing Copilot?

Bing Copilot is Microsoft’s AI assistant built into Bing Search. It uses Microsoft’s proprietary Copilot AI model to interpret your query and synthesize an answer. Copilot blends the best of traditional and generative search, so that instead of sifting through links you get an easy-to-digest answer with key points. For example, a Copilot answer might begin with a concise summary, optionally include a supporting image or video, and then list the sources used.

Bing Copilot’s answer to a query (here “Roman Empire”) appears as a summary card with text, an image, and a list of numbered source links.

Each Copilot answer is a generated summary that comes with clearly cited references. This makes Copilot powerful for research: you get the key answer up front and can click any source to verify it. SerpApi’s Bing Copilot API captures all of this in a structured JSON format. In the JSON response you’ll typically see a header (the main answer text), an optional header_video (if a video result is featured), and an array of text_blocks – where each block might be a paragraph, heading, list, code snippet, or table. At the end of the JSON is a references list containing the source URLs and titles. This mirrors Bing’s behavior: Copilot always provides links to the web pages used to generate the answer.

Why scrape Bing Copilot?

Bing Copilot’s answers are filled with actionable data that businesses and developers can use. For example:

• Generative Search Optimization (GEO) – Ensure your content is accurately represented in AI answers. Scraping Copilot lets you see how your pages are cited and summarized by the AI.
• SEO & Content Strategy – Identify which sites and keywords Bing’s AI favors. By extracting source links and answer content, you can spot content gaps and refine what topics you cover.
• Competitive Intelligence & Brand Monitoring – Monitor how often your brand or competitors appear in Copilot answers. This helps you track your presence (or absence) in AI-driven results.
• Market Research – Aggregate Copilot’s synthesized answers across queries to uncover trends. The AI’s summaries often surface insights spanning multiple sites.
• AI/LLM Training Data – Use Copilot outputs (with citations) as high-quality, real-world examples. The structured answer + source links make excellent training or evaluation data for other AI models.
• Reputation & Insight Tracking – See how your products, executives, or topics are described by Bing’s AI. You can spot misleading or negative summaries early by monitoring the Copilot output.

Scraping Bing Copilot answers programmatically lets you tap all these use cases. You can also run analytics, dashboards, or feeds without manual copy-paste.

Why use SerpApi for Copilot scraping?

You could try to build your own Bing Copilot scraper, but it would be a lot of work. Bing’s Copilot interface uses dynamic HTML and may change over time. It also has rate limits and anti-bot defenses. SerpApi handles all that complexity for you. When you use the Bing Copilot API:

• Accurate Parsing – The API returns the Copilot answer already parsed into JSON (with separate text blocks and references), so you don’t have to reverse-engineer Bing’s HTML.
• Anti-Block Infrastructure – SerpApi automatically rotates IPs and solves captchas behind the scenes, so you won’t get blocked by Bing’s protection systems.
• Up-to-Date Support – The SerpApi team continuously monitors Bing for interface changes. When Copilot is updated, SerpApi updates the parser so your API calls keep working.
• Scalability – You can run hundreds or thousands of Copilot queries per day without managing proxies or servers. SerpApi handles the scaling so you can focus on your data.

Querying the Bing Copilot API

Getting started is easy. Sign up at SerpApi and grab your API key. Then you can query the Copilot API by sending a GET request to SerpApi with engine=bing_copilot and your search query q. You can also test it interactively in the SerpApi Playground by selecting the “Bing Copilot” engine.

Python example

from serpapi import GoogleSearch

params = {
    "engine": "bing_copilot",
    "q": "how many moons does Jupiter have",
    "api_key": "YOUR_SERPAPI_KEY"
}
search = GoogleSearch(params)
result = search.get_dict()
print(result["header"])
# e.g. "Jupiter has 95 known moons..."

This will print the Copilot’s answer to the query and the returned result dictionary will include all text blocks and references.

JavaScript (Node.js) example

const SerpApi = require('google-search-results-nodejs');
const search = new SerpApi.GoogleSearch("YOUR_SERPAPI_KEY");

search.json({
  engine: "bing_copilot",
  q: "python bubble sort example"
}, (data) => {
  console.log(data.header);        // main answer text
  console.log(data.text_blocks);   // array of paragraphs, lists, code, etc.
  console.log(data.references);    // array of source links
});

With the JSON result in hand, you can parse out exactly what you need. For example, you could iterate over result["text_blocks"] to extract all list items or all code snippets. Because Copilot’s answer is broken into blocks (paragraphs, lists, tables, code, etc.), it’s easy to transform or filter parts of the answer programmatically.

Try It Out!

The new Bing Copilot API is available now to all SerpApi users. To get started, create a free SerpApi account (the free tier includes 250 queries/month) and find your API key. Then visit the Playground or use the endpoints above to begin scraping Copilot answers. You can find API documentation here. Whether you’re optimizing content for AI search, monitoring brand mentions in AI summaries, or building new data insights from generative answers, the Bing Copilot API makes it simple.