Telegram Public Channel Scraper

Collect messages from public Telegram channel previews at https://t.me/s/<channel>. No Telegram API key is used. Only information actually present in the public preview is emitted; unavailable media and counters are omitted.

Input

• channels (required): one or more public usernames, @usernames, or t.me channel/preview links
• maxMessages: maximum saved messages per channel, from 1 to 500; defaults to 50
• proxyConfiguration: optional preconfigured proxy settings

{"channels":["telegram"],"maxMessages":10}

Output

Each row contains messageId, messageUrl, channelUsername, channelTitle, channelUrl, the source datetime, channel-local rank and channelRank, sourcePageUrl, source, provenance, and scrapedAt. Optional source fields include:

• text, channelDescription
• subscriberCount, views, replyCount, with their original *Display strings
• mediaType and mediaTypes: photo, video, voice, audio, document, sticker or poll; a mixed message has mediaType: "mixed"
• mediaUrls: deduplicated photo, video or audio source URLs found in the preview; media files are not downloaded
• mediaThumbnails: separately reported video poster/thumbnail URLs
• mediaDurationDisplay: source duration labels, without estimating a numeric duration
• documentName and documentMeta: visible filename and accompanying source metadata; no file size is inferred from rounded labels
• linkPreviewTitle, linkPreviewUrl, linkPreviewDescription, linkPreviewDomain (the validated link hostname), and linkPreviewSiteName (the displayed site label)
• isForwarded: true, forwardedFrom, and forwardedFromUrl when a forwarding element is present

Missing counters and forwarding information remain absent. Explicit zero counters are retained. Abbreviated counters such as 1.2K are normalized to 1200, while the original display value is retained; this is a rounded display count, not a claim of exact underlying precision. Invalid/negative counters are not converted into positive facts. Unknown or protected media URLs are never fabricated.