Usage examples

# Clean markdown - great for reading or feeding to LLMs
brightdata scrape https://news.ycombinator.com

# Save to file
brightdata scrape https://docs.python.org/3/tutorial/ -o python-tutorial.md

# Get raw HTML for parsing
brightdata scrape https://example.com -f html -o page.html

# See Amazon prices as a US customer
brightdata scrape https://amazon.com/dp/B09V3KXJPB --country us

# Scrape a German news site from Germany
brightdata scrape https://spiegel.de --country de

# Mobile user agent for mobile-optimized pages
brightdata scrape https://example.com --mobile

# Full-page screenshot
brightdata scrape https://example.com -f screenshot -o homepage.png

# Screenshot from a specific country
brightdata scrape https://amazon.co.uk -f screenshot --country gb -o uk-amazon.png

# Google search with formatted table output
brightdata search "best web scraping tools 2025"

# Get raw JSON for processing
brightdata search "typescript best practices" --json

# Pretty-print for inspection
brightdata search "AI startups" --pretty

# Local restaurant search from Germany, in German
brightdata search "restaurants berlin" --country de --language de

# News-only results
brightdata search "AI regulation 2025" --type news

# Shopping results
brightdata search "wireless headphones" --type shopping

# Image search
brightdata search "mountain landscape wallpaper" --type images

# First page (default)
brightdata search "web scraping tutorials"

# Second page
brightdata search "web scraping tutorials" --page 1

# Third page
brightdata search "web scraping tutorials" --page 2

# Open a page and read its content
brightdata browser open https://example.com
brightdata browser snapshot

# Compact snapshot for AI agents (70-90% fewer tokens)
brightdata browser snapshot --compact

# Scope to a specific section
brightdata browser snapshot --selector "main"

# Open a page
brightdata browser open https://example.com

# Read the page structure
brightdata browser snapshot --compact

# Click, type, and submit using refs from the snapshot
brightdata browser click e3
brightdata browser type e5 "search query" --submit

# Get updated snapshot after interaction
brightdata browser snapshot --compact

# Take a screenshot for visual verification
brightdata browser screenshot ./result.png

# Clean up
brightdata browser close

# Open the same page from different countries
brightdata browser open https://amazon.com --session us --country us
brightdata browser open https://amazon.com --session de --country de

# Capture both snapshots
brightdata browser snapshot --session us --json > us.json
brightdata browser snapshot --session de --json > de.json

# Close all sessions
brightdata browser close --all

# Navigate to a page that requires JavaScript
brightdata browser open https://example.com/dashboard

# Get text content from specific elements
brightdata browser get text "#total-revenue"
brightdata browser get text ".summary-table"

# Get HTML for parsing
brightdata browser get html ".product-grid"

# Check network requests
brightdata browser network

# Search by site or data type: returns dataset_id, name, domain and collection_methods
brightdata pipelines search amazon reviews

# List the scraper's collection methods
brightdata pipelines info gd_le8e811kzy4ggddlq

# Show one method's input fields, an example input and a ready-to-run command
brightdata pipelines info gd_le8e811kzy4ggddlq collect_by_url

# Amazon product details, one record per URL
brightdata pipelines run gd_l7q7dkf244hwjntr0 collect_by_url "https://www.amazon.com/dp/B0FQFB8FMG"

# Amazon reviews, up to 20 per product
brightdata pipelines run gd_le8e811kzy4ggddlq collect_by_url "https://www.amazon.com/dp/B0FQFB8FMG" --max_reviews 20

# Walmart products by keyword, capped at 5 records (every record is billed)
brightdata pipelines run gd_l95fol7l1ru6rlo116 discover_by_keyword \
  --keyword "wireless headphones" --domain "https://www.walmart.com/" --limit-per-input 5

# Export as CSV
brightdata pipelines run gd_l7q7dkf244hwjntr0 collect_by_url "https://www.amazon.com/dp/B0FQFB8FMG" --format csv -o product.csv

# LinkedIn person
brightdata pipelines run gd_l1viktl72bvl7bjuj0 collect_by_url "https://www.linkedin.com/in/satyanadella"

# LinkedIn company
brightdata pipelines run gd_l1vikfnt1wgvvqz95w collect_by_url "https://www.linkedin.com/company/bright-data"

# Two Instagram profiles in one run, one record per URL
brightdata pipelines run gd_l1vikfch901nx3by4 collect_by_url \
  "https://www.instagram.com/australiazoo" "https://www.instagram.com/zoobarcelona"

# Google Maps reviews from the last 7 days
brightdata pipelines run gd_luzfs1dn2oa0teb81 collect_by_url "<google_maps_place_url>" --days_limit 7

# YouTube comments, top 50
brightdata pipelines run gd_lk9q0ew71spt1mxywf collect_by_url "https://www.youtube.com/watch?v=dQw4w9WgXcQ" --num_of_comments 50

# Facebook company reviews, 25 reviews
brightdata pipelines run gd_m0dtqpiu1mbcyc2g86 collect_by_url "https://www.facebook.com/KFC/reviews/" --num_of_reviews 25

# Search Google, extract the first URL, then scrape it
brightdata search "best python frameworks 2025" --json \
  | jq -r '.organic[0].link' \
  | xargs brightdata scrape

# Pipe markdown output to a terminal reader
brightdata scrape https://docs.github.com | glow -

# Or use less
brightdata scrape https://docs.github.com | less

# Amazon product data to CSV
brightdata pipelines run gd_l7q7dkf244hwjntr0 collect_by_url "https://www.amazon.com/dp/B0FQFB8FMG" --format csv > product.csv

# LinkedIn job listing to CSV
brightdata pipelines run gd_lpfll7v5hcqtkxl6l collect_by_url \
  "https://www.linkedin.com/jobs/view/software-engineer-at-pave-4310512612/" --format csv -o jobs.csv

# Get just the titles and prices from an Amazon keyword search
brightdata pipelines run gd_l7q7dkf244hwjntr0 discover_by_keyword --keyword "laptop" --limit-per-input 5 \
  | jq '[.[] | {title, price: .final_price}]'

# Get just organic result URLs from search
brightdata search "web scraping" --json | jq -r '.organic[].link'

# Start a run without waiting; --async prints the snapshot ID
SNAPSHOT_ID=$(brightdata pipelines run <dataset_id> <method> "<url>" --async --json | jq -r '.snapshot_id')

# Do other work...

# Check back later, then fetch the records
brightdata pipelines status $SNAPSHOT_ID --wait
brightdata pipelines results $SNAPSHOT_ID --pretty

# Quick balance check
brightdata budget

# Detailed balance with pending charges
brightdata budget balance

# Cost breakdown by zone
brightdata budget zones

# Specific zone in a date range
brightdata budget zone cli_unlocker --from 2024-01-01T00:00:00 --to 2024-02-01T00:00:00

# View current config
brightdata config

# Set default output to JSON
brightdata config set default_format json

# Use a custom zone for scraping
brightdata config set default_zone_unlocker my_custom_zone

# Override SERP zone
brightdata config set default_zone_serp my_serp_zone

# Interactive - choose agent and scope
brightdata add mcp

# Add to Claude Code globally
brightdata add mcp --agent claude-code --global

# Add to multiple agents at once
brightdata add mcp --agent claude-code,cursor --project

# Add to Codex
brightdata add mcp --agent codex --global

# Interactive picker - choose skills and target agents
brightdata skill add

# Install the scraping skill into your agent
brightdata skill add scrape

# Install search capabilities
brightdata skill add search

# See all available skills
brightdata skill list

# Use in CI/CD without login
BRIGHTDATA_API_KEY=your_key brightdata scrape https://example.com

# Override timeout for large pipeline jobs
BRIGHTDATA_POLLING_TIMEOUT=1200 brightdata pipelines run <dataset_id> <method> "<url>"