Telegram Channel Scraper - Public Posts and Channel Info

Scrape public Telegram channels without a login, API key or phone number. For each channel you get the messages (text, date, views, reactions, media, link previews, forwards, links, hashtags) and, if you want, the channel profile (title, description, subscribers, photo, verified flag). Search for a keyword inside a channel or go back in time up to a date.

What data can you extract?

Message rows (entry_type: "message")

• Message ID, direct message URL, channel name and title
• Publication date (ISO 8601, UTC), edited flag, author signature
• Plain text and HTML text, external links, hashtags, @mentions
• View count (as shown by Telegram, rounded for large numbers)
• Reactions with count (emoji or custom emoji ID, paid star reactions)
• Media: photos, videos, round videos, voice messages, documents, stickers, polls, with URLs where Telegram publishes them
• Link preview card (URL, site name, title, description, image)
• Forwarded-from source and reply-to reference

Channel rows (entry_type: "channel", optional)

• Title, description (plain and HTML), links in the description
• Subscriber count (rounded as shown by Telegram), counts of photos, videos, files and links
• Verified flag, channel photo URL

Telegram does not show reply counts or forward counts on the public web preview, so those fields are not available.

Use with AI Agents (MCP)

Connect this actor to any MCP-compatible AI client, including Claude Desktop, Claude.ai, Cursor, VS Code, LangChain and LlamaIndex.

{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.apify.com?tools=santamaria-automations/telegram-scraper",
        "--header",
        "Authorization: Bearer <APIFY_TOKEN>"
      ]
    }
  }
}

Example prompt: "Use telegram-scraper to get the last 50 posts of the durov channel and summarise the topics with view counts."

How to use

{
  "channels": ["telegram", "@durov", "https://t.me/bbcnews"],
  "maxMessages": 100,
  "searchQuery": "",
  "includeChannelInfo": true
}

An empty input {} also works and runs the defaults (the telegram channel, 20 messages, plus the channel profile).

Input parameters

Parameter	Type	Default	Description
channels	array	["telegram"]	Channel names: name, @name or a t.me link
maxMessages	integer	20	Maximum messages per channel, newest first (1-10,000)
searchQuery	string	empty	Only messages containing this word or phrase, searched inside each listed channel
dateFrom	string	empty	Stop at messages older than this date (YYYY-MM-DD)
includeChannelInfo	boolean	true	Add one profile row per channel

Limits

• Only channels with a public web preview work. Private channels, invite links (t.me/+...) and channels that disabled the preview are reported and skipped.
• Telegram offers no keyword search across all channels without a login. searchQuery searches inside the channels you list.
• Large counters (views, subscribers, reactions) are rounded by Telegram, for example 1.59M.

What you get

Message rows and channel profile rows share one table; fields marked "(channel rows)" are only filled for profile rows, message fields are empty (null) on profile rows.

Field	Type	Description
entry_type	string	message or channel (channel profile row, only if includeChannelInfo)
channel	string	Channel username without @
channel_title	string	Channel display name
channel_url	string	Channel URL
id	integer	Message ID, unique within the channel (message rows)
message_url	string	Direct link to the message
date	string	Publication time (ISO 8601, UTC)
text	string	Plain text of the message
text_html	string	Message text as HTML (formatting and links kept)
views_text	string	View count as shown by Telegram
author_signature	string	Post author signature if the channel shows one
edited	boolean	True if the message was edited
forwarded_from	string	Original source of a forwarded message
forwarded_from_url	string	Link to the original source
reply_to_url	string	Message this one replies to
reply_to_message_id	integer	ID of the replied message
reply_to_text	string	Preview of the replied message
has_media	boolean	True if the message has media
media_type	string	Type of the first media item: photo, video, round_video, voice, document, sticker, poll, location
media	array	All media items with type, url, thumbnail_url, title, duration, size_text, poll_question, poll_options
link_preview	object	Link preview card: url, site_name, title, description, image_url
links	array	External links in the message text
hashtags	array	Hashtags in the message text
mentions	array	@mentions in the message text
reactions	array	Reactions with emoji, custom_emoji_id, paid, count, count_text
search_query	string	Keyword used for the in-channel search
title	string	Channel title (channel rows)
description	string	Channel description (channel rows)
description_html	string	Channel description as HTML (channel rows)
subscribers	integer	Subscriber count, rounded as shown by Telegram (channel rows)
subscribers_text	string	Subscriber count as shown (channel rows)
photos_count	integer	Number of photos in the channel (channel rows)
videos_count	integer	Number of videos in the channel (channel rows)
files_count	integer	Number of files in the channel (channel rows)
links_count	integer	Number of links in the channel (channel rows)
verified	boolean	Verified channel (channel rows)
photo_url	string	Channel photo URL (channel rows)
scraped_at	string	When the row was scraped

Output example

{
  "entry_type": "message",
  "channel": "durov",
  "channel_title": "Pavel Durov",
  "id": 548,
  "message_url": "https://t.me/durov/548",
  "date": "2026-09-11T16:04:02+00:00",
  "text": "Telegram has become the sponsor of Codeforces ...",
  "views": 3550000,
  "views_text": "3.55M",
  "author_signature": "Pavel Durov",
  "has_media": false,
  "link_preview": {"url": "https://codeforces.com/blog/entry/156620", "site_name": "Codeforces", "title": "Telegram Returns as Title Sponsor of Codeforces!"},
  "links": ["https://codeforces.com/blog/entry/156620"],
  "reactions": [{"emoji": null, "custom_emoji_id": "5399847211989246390", "paid": false, "count": 36300, "count_text": "36.3K"}],
  "scraped_at": "2026-10-21T15:39:33Z"
}

Fields Telegram does not publish for a message are null (or an empty list).

Pricing

Pay per result. You are charged one start event per run and one event for every row saved to the dataset.

Event	Price	When
telegram-start	$0.005	Once per run, right before the first row is saved
telegram-message-result	$0.0012	Per message saved ($1.20 per 1,000 messages)
telegram-channel-result	$0.002	Per channel profile row saved (only if includeChannelInfo is on)

Runs without results cost nothing. Duplicates are never saved or charged twice, and a restarted run does not charge rows it already saved.

Performance

• Speed: about 20 messages per page and one page per 1-2 seconds, so roughly 1,000 messages per minute
• Transfer: about 1.2 MB per 1,000 messages

Tips

• Use dateFrom for daily or weekly monitoring so you only collect new posts
• Set includeChannelInfo to false if you only need the messages
• Combine several channels in one run; the start event is charged once

Related actors

• Instagram Scraper - posts and profiles from Instagram.
• TikTok Search Scraper - videos and creators from TikTok.
• Reddit Scraper - posts and comments from Reddit.
• YouTube Channel Scraper - channel data from YouTube.

Data protection

This actor reads only information that is publicly visible on the Telegram web preview of public channels. Messages, author signatures, forwarded-from names and linked profiles can contain personal data of named individuals. You are responsible for having a lawful basis (for example legitimate interest under Article 6 GDPR) for the way you store and use that data, and for complying with GDPR, national data-protection laws, unfair competition acts and Telegram's terms. Review your use case with a qualified data protection officer before large-scale processing or outreach. Do not use scraped data to contact people who have not agreed to it where the law requires consent.