Skip to main content
Send any public URL, get the page back as Markdown, plain text, raw HTML and JSON-LD in a single record. Fetch API (formerly Crawl API) is the Scraper API equivalent of Web Unlocker API: Bright Data handles proxies, CAPTCHAs and blocking, and the Fetch API returns the page content ready for an LLM or a parser.
New to Bright Data? Create a free account and get 5,000 free credits every month, no credit card required. See the free tier.

​
How it works

You send one or more URLs to the Bright Data Fetch API. Bright Data fetches each page and returns one record per URL with every output format bundled together.
Every request uses the Fetch API dataset_id, gd_m6gjtfmeh43we6cqc, and returns results in NDJSON (default), JSON, JSON lines, CSV and Parquet. See output formats.

​
What the response looks like

A single record returned by the Bright Data Fetch API for https://example.com on October 19, 2026. Run the request yourself in the quickstart.
The page_html value is trimmed here. The example page has no JSON-LD, so the record has no ld_json field.

​
Which formats does Fetch API return?

Markdown

markdown keeps headings, lists and links, ready for an LLM prompt or a RAG index.

Plain text

html2text strips every HTML tag and keeps only the visible text.

Raw HTML

page_html is the complete HTML source, including tags, scripts and styles.

JSON-LD

ld_json holds the page’s structured schema.org metadata, when the page has any.
Each record also carries url, final_url (the URL after redirects), page_title and country.

​
Which dataset ID to use

The Fetch API has one endpoint, collect by URL. Pass the ID as a query parameter on every request.

​
Request methods

The Bright Data Scraper API supports two request methods. Choose based on your volume and latency needs. Learn more in Understanding sync vs. async requests. Both endpoints take an object with an input array, plus optional limit_per_input and custom_output_fields, or a deliver block on /trigger. A bare JSON array of inputs is also accepted, and the examples on this page use it. See Request body.

​
Capabilities and limits

​
Common questions

​
How is Fetch API different from Web Unlocker API?

Both unblock any public web page. They differ in how you call them and what you get back:
  • Web Unlocker API is called at /request with a Web Unlocker zone and returns the target page’s raw response, as-is or in a JSON envelope.
  • Fetch API is called at /datasets/v3/scrape or /datasets/v3/trigger with a dataset_id and needs no zone. Each URL comes back as one record holding Markdown, plain text, raw HTML and JSON-LD together, and works with Scraper API batch jobs and delivery options.

​
Is the page fetched in real time?

Yes. Each request fetches the live page. There is no cached copy. A synchronous request waits up to 1 minute, then returns HTTP 202 with a snapshot_id and continues asynchronously. Processing time varies with target-site conditions, so the Bright Data Fetch API does not publish a fixed response time.

​
Which websites does Fetch API work on?

The Fetch API accepts any public URL on any domain. It is not tied to one platform the way the Instagram or Amazon scrapers are, and it does not parse site-specific fields such as prices or follower counts. Use a platform scraper when you need those fields as structured JSON.

​
Next steps

Quickstart

Fetch your first page as Markdown in 5 minutes.

Endpoints

Country targeting, field selection and batch requests.

API reference

Endpoint specs, parameters, and response schemas.