supadata 1.6.0

The official Python SDK for Supadata - extract web media data with ease

Supadata Python SDK

The official Python SDK for Supadata.

Get your free API key at supadata.ai and start scraping data in minutes.

Installation

pip install supadata

Usage

Initialization

from supadata import Supadata, SupadataError

# Initialize the client
supadata = Supadata(api_key="YOUR_API_KEY")

Metadata

# Get media metadata from any supported platform (YouTube, TikTok, Instagram, Twitter)
metadata = supadata.metadata(url="https://www.youtube.com/watch?v=dQw4w9WgXcQ")
print(metadata)

Transcripts

# Get transcript from any supported platform (YouTube, TikTok, Instagram, Twitter, file URLs)
transcript = supadata.transcript(
    url="https://x.com/SpaceX/status/1481651037291225113",
    lang="en",  # Optional: preferred language
    text=True,  # Optional: return plain text instead of timestamped chunks
    mode="auto"  # Optional: "native", "auto", or "generate"
)

# For immediate results
if hasattr(transcript, 'content'):
    print(f"Transcript: {transcript.content}")
    print(f"Language: {transcript.lang}")
else:
    # For async processing (large files)
    print(f"Processing started with job ID: {transcript.job_id}")
    # Poll for results using existing batch.get_batch_results method

Extract

# Extract structured data from video using AI
job = supadata.extract(
    url="https://www.youtube.com/watch?v=dQw4w9WgXcQ",
    prompt="Extract the main topics and key takeaways",  # Optional if schema is provided
    schema={  # Optional if prompt is provided
        "type": "object",
        "properties": {
            "topics": {"type": "array", "items": {"type": "string"}},
            "summary": {"type": "string"},
        },
        "required": ["topics", "summary"],
    },
)
print(f"Started extract job: {job.job_id}")

# Get extract results (poll until status is 'completed' or 'failed')
result = supadata.extract.get_results(job_id=job.job_id)
print(f"Status: {result.status}")
print(f"Data: {result.data}")

YouTube

# Translate YouTube transcript to Spanish
translated = supadata.youtube.translate(
    video_id="dQw4w9WgXcQ",
    lang="es"
)
print(f"Got translated transcript in {translated.lang}")

# Get Channel Metadata
channel = supadata.youtube.channel(id="https://youtube.com/@RickAstleyVEVO") # can be url, channel id, handle
print(f"Channel: {channel}")

# Get video IDs from a YouTube channel
channel_videos = supadata.youtube.channel.videos(
    id="RickAstleyVEVO",  # can be url, channel id, or handle
    type="all",  # 'all', 'video', 'short', or 'live'
    limit=50
)
print(f"Regular videos: {channel_videos.video_ids}")
print(f"Shorts: {channel_videos.short_ids}")
print(f"Live: {channel_videos.live_ids}")

# Get Playlist metadata
playlist = supadata.youtube.playlist(id="PLlaN88a7y2_plecYoJxvRFTLHVbIVAOoc") # can be url or playlist id
print(f"Playlist: {playlist}")

# Get video IDs from a YouTube playlist
playlist_videos = supadata.youtube.playlist.videos(
    id="https://www.youtube.com/playlist?list=PLlaN88a7y2_plecYoJxvRFTLHVbIVAOoc",  # can be url or playlist id
    limit=50
)
print(f"Regular videos: {playlist_videos.video_ids}")
print(f"Shorts: {playlist_videos.short_ids}")
print(f"Live: {playlist_videos.live_ids}")

# Search YouTube videos
search_results = supadata.youtube.search(
    query="Never Gonna Give You Up",
    upload_date="all",  # "all", "hour", "today", "week", "month", "year"
    type="video",       # "all", "video", "channel", "playlist", "movie"
    duration="all",     # "all", "short", "medium", "long"
    sort_by="relevance", # "relevance", "rating", "date", "views"
    features=["hd", "subtitles"],  # Optional: filter by video features
    limit=10            # Optional: number of results (1-5000)
)
print(f"Found {search_results.total_results} total results")
print(f"Query: {search_results.query}")
for result in search_results.results:
    print(f"Video: {result.title} by {result.channel['name']}")
    print(f"  ID: {result.id}")
    print(f"  Duration: {result.duration}s")
    print(f"  Views: {result.view_count}")


# Batch Operations
transcript_batch_job = supadata.youtube.transcript.batch(
    video_ids=["dQw4w9WgXcQ", "xvFZjo5PgG0"],
    # playlist_id="PLlaN88a7y2_plecYoJxvRFTLHVbIVAOoc", # alternatively
    # channel_id="UC_9-kyTW8ZkZNDHQJ6FgpwQ", # alternatively
    lang="en",  # Optional: specify preferred transcript language
    limit=100   # Optional: limit for playlist/channel
)
print(f"Started transcript batch job: {transcript_batch_job.job_id}")

# Start a batch job to get video metadata for a playlist
video_batch_job = supadata.youtube.video.batch(
    playlist_id="PLlaN88a7y2_plecYoJxvRFTLHVbIVAOoc",
    limit=50
)
print(f"Started video metadata batch job: {video_batch_job.job_id}")

# Get the results of a batch job (poll until status is 'completed' or 'failed')
batch_results = supadata.youtube.batch.get_batch_results(job_id=transcript_batch_job.job_id)
print(f"Job status: {batch_results.status}")
print(f"Stats: {batch_results.stats.succeeded}/{batch_results.stats.total} videos processed")
print(f"First result: {batch_results.results[0].video_id if batch_results.results else 'No results yet'}")

Web

# Scrape web content
web_content = supadata.web.scrape("https://supadata.ai")
print(f"Page title: {web_content.name}")
print(f"Page content: {web_content.content}")

# Map website URLs
site_map = supadata.web.map("https://supadata.ai")
print(f"Found {len(site_map.urls)} URLs")

# Start a crawl job
crawl_job = supadata.web.crawl(
    url="https://supadata.ai",
    limit=100  # Optional: limit the number of pages to crawl
)
print(f"Started crawl job: {crawl_job.job_id}")

# Get crawl results
# This automatically handles pagination and returns all pages
try:
    pages = supadata.web.get_crawl_results(job_id=crawl_job.job_id)
    for page in pages:
        print(f"Crawled page: {page.url}")
        print(f"Page title: {page.name}")
        print(f"Content: {page.content}")
except SupadataError as e:
    print(f"Crawl job failed: {e}")

Error Handling

The SDK uses custom SupadataError exceptions that provide structured error information:

from supadata.errors import SupadataError

try:
    metadata = supadata.metadata(url="https://www.youtube.com/watch?v=INVALID_ID")
except SupadataError as error:
    print(f"Error code: {error.error}")
    print(f"Error message: {error.message}")
    print(f"Error details: {error.details}")
    if error.documentation_url:
        print(f"Documentation: {error.documentation_url}")

API Reference

See the Documentation for more details on all possible parameters and options.

License

MIT