Telegram Channel Scraper

What does Telegram Channel Scraper do?

Telegram Channel Scraper reads public Telegram channels and turns their messages into a table. Give it channel names such as durov or links such as https://t.me/telegram, and you get one row per message with its text, date, views, reactions, photos and videos, links, hashtags, forwards and replies, plus one row with the channel's own details: title, description and number of subscribers. Export to Excel, CSV or JSON, or send it on to another app.

There is nothing to set up. No Telegram account, no phone number, no API key and no bot token: the tool reads the public web preview that Telegram shows to anyone at t.me/s/channelname.

• Every message of a channel, newest first, as far back as you ask
• Filter by date: the last 24 hours, a month in the past, or any period in between
• Filter by keyword or hashtag, using Telegram's own channel search
• Many channels in one run, each message once
• Text as plain text and as HTML, with the links, hashtags and mentions listed
• Views, reactions with their counts, forwarded-from, reply-to, edited mark and author signature
• Photos, videos and link previews with their addresses, video lengths and thumbnails
• Channels that do not exist or have no public preview give a clear error row and never stop the run

What data do you get?

Each message is one row:

Field	Example
Channel	durov and the channel's title
Message	https://t.me/durov/531, number 531
Date	2026-07-16T20:45:25.000Z
Text	the message as plain text, and as HTML (textHtml)
Views	5830000 (Telegram shows 5.83M)
Reactions	[{"emoji": "👍", "count": 54600, "paid": false}]
Forwarded from / Reply to	the source's name and link, the replied message's number, link and text
Author	the signature on the post, for channels that sign their posts
Edited	true or false
Links, hashtags, mentions	the addresses, #tags (without the sign) and @names in the text
Media	photos and videos with address, thumbnail and length in seconds; voice messages, files, stickers, polls and locations where the preview shows them
Link preview	the card Telegram builds for a link: address, site, title, description, image
Buttons	link buttons under the post
Keyword	the keyword whose search found the message

The channel row (type: "channel") has the channel's title, description, subscribers, the counts of photos, videos, files and links that Telegram shows, the verified mark and the profile picture. It comes first for each channel and counts as one result. Turn it off with Include a row with the channel's details, or ask for nothing else with Channel details only: a quick way to compare the size of many channels.

How to scrape a Telegram channel

1. Under Channels, add the channels: user names, @names or links, one per line.
2. Optionally set Maximum messages per channel, Posted from / until or Keywords.
3. Press Start.
4. Export the results as Excel or CSV, or connect them to Google Sheets, Slack, Make or Zapier.

How much does it cost to scrape Telegram?

You pay for results, not for time: $0.50 per 1,000 messages. Apify subscribers on the Silver plan pay $0.45 and on the Gold plan $0.40 per 1,000 messages. Each run also has a start fee of $0.00005.

Messages returned	Cost
100	$0.05
1,000	$0.50
10,000	$5.00
100,000	$50.00

Apify's free plan includes $5 of credit each month, which covers about 10,000 messages. A channel row and a row for a channel that could not be read each count as one result. Messages left out by your filters are not charged. Set a maximum cost for the run in Apify Console and the tool stops there and keeps what it has read.

Input

The only required field is Channels. The other options:

• Maximum messages per channel: keeps the newest messages of each channel. Telegram shows 20 messages per page and answers Apify's servers in a few seconds per page: in a test on Apify, 1,864 messages from three channels took 6.5 minutes.
• Posted from and Posted until: a date, or a period such as 7 days or 24 hours. A plain end date includes the whole day. Reading a period far in the past is fast: the tool jumps to the end date instead of reading every newer page.
• Keywords: words, phrases or hashtags such as bitcoin or #news. Telegram's own channel search finds the messages, so a match is what Telegram counts as one. Keywords are searched one after the other (up to 25), and a message found by several is returned once.
• Include a row with the channel's details and Channel details only.
• Time limit per request and User-Agent under More options.

Output example

{
    "type": "message",
    "channel": "example_news",
    "channelTitle": "Example News",
    "messageId": 4812,
    "url": "https://t.me/example_news/4812",
    "date": "2026-10-22T17:40:10.000Z",
    "text": "Central bank holds rates (Score: 155 in 2 hours)\n\nLink: https://example.com/rates",
    "textHtml": "<b>Central bank holds rates</b> (Score: 155 in 2 hours)<br><br>Link: <a href=\"https://example.com/rates\">https://example.com/rates</a>",
    "views": 2110,
    "reactions": [{ "emoji": "👍", "customEmojiId": null, "count": 36, "paid": false }],
    "forwardedFrom": null,
    "replyTo": null,
    "author": null,
    "edited": false,
    "links": ["https://example.com/rates"],
    "hashtags": ["economy"],
    "mentions": [],
    "media": [{ "type": "photo", "url": "https://cdn4.telesco.pe/file/example.jpg", "thumbnail": null, "duration": null }],
    "linkPreview": { "url": "https://example.com/rates", "siteName": "Example", "title": "Central bank holds rates", "description": "The bank kept its rate at 3.5 %.", "imageUrl": null },
    "buttons": [],
    "keyword": null
}

A channel that cannot be read gives one row with an error: NO_PUBLIC_PREVIEW, CHANNEL_NOT_FOUND, INVALID_INPUT (a private invite link or something that is not a channel), RATE_LIMITED, ROBOTS_DISALLOWED or a network error. If a page fails in the middle of a channel, the messages read so far are kept and the error row says how many.

What can you use it for?

• Monitor news, announcement and crypto channels, and get new posts into Slack, e-mail or your own app on a schedule
• Collect posts for sentiment analysis, research and AI tools
• Track which topics, hashtags and links a channel pushes, and how many views and reactions they get
• Compare the size and activity of channels in a niche
• Archive a channel before posts are deleted

Schedule the run in Apify Console, for example every hour with Posted from set to 1 hour, to get only the new messages each time.

Use it from the API

curl -X POST "https://api.apify.com/v2/acts/cyber1nj3ct0r~telegram-channel-scraper/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"channels": ["durov", "https://t.me/s/telegram"], "maxMessagesPerChannel": 200, "postedFrom": "7 days"}'

The same works from Python, JavaScript, Make, Zapier and n8n through Apify's integrations.

FAQ

Which channels can it read?

Public channels whose preview opens at t.me/s/channelname in a browser without logging in. Private channels, invite links, groups, bots and people's accounts have no such page, and the tool says so in an error row. A channel whose owner turned the web preview off cannot be read either. Telegram gives the same answer for a name nobody has taken, so the error also covers a misspelled name.

Do I need a Telegram account or an API key?

No. The tool reads the same page a visitor sees. It identifies itself in its User-Agent, follows the site's robots.txt, asks for about one page per second, waits when Telegram asks it to slow down, and uses no logins, cookies or proxies.

How far back can it go?

As far as the channel's public preview does. In tests it read @durov and @telegram back to their first message. Set Maximum messages per channel and the dates to limit the cost.

Why are the view counts rounded?

Telegram shows 5.83M or 2.1K, not exact numbers, and so do the subscribers and reactions. The tool turns them into numbers (5830000) but cannot know more than the page says.

What does the media type unsupported mean?

For some posts the preview hides part of the content behind a "Please open Telegram to view this post" box. Voice messages, files, polls and some other items can be hidden this way. The tool can only return what the preview shows, so the row then has a media entry of type unsupported with that note, and nothing else about the hidden part. Albums sent together appear as one row with all their photos, numbered with the first message.

Do the photo and video links last?

They are Telegram's own file addresses and can stop working after a while, and the videos carry a token. Download what you need soon after the run. Videos that are too big for the preview have a thumbnail and a length but no file address.

Does it get comments, member lists or private chats?

No. It returns the posts of public channels as the preview shows them: no comments under posts, no list of members and nothing from private chats.

Is it allowed to scrape Telegram?

The tool reads only pages that anyone can open without logging in. Messages can still contain personal data and are copyrighted by their authors: you are responsible for how you store, use and share them, under the laws that apply to you and Telegram's terms.

Something does not work

Open an issue on the Issues tab with the channel you tried. Telegram changes its pages from time to time, and every report is read and answered.