Telegram Channel Scraper - Posts & Views, No Login

Scrape posts from public Telegram channels with no login, no phone number and no API keys, and get clean JSON rows with text, date, view count as a number, media links, reactions, forwards and channel info with the exact subscriber count.

Use it for news and crypto monitoring, research archives or AI pipelines: the smallest run is one channel name, such as durov, and two channels with 20 posts each finish in about 5 seconds for about $0.08 on the Free plan.

Table of contents

• What it does
• Quick start
• Input
• Output
• Pricing
• Use cases
• Scheduling and monitoring
• Switching from another Telegram scraper
• API, export and integrations
• Limits and partial results
• Troubleshooting
• FAQ
• Learn more
• Support

What it does

The Actor reads the public web preview of a Telegram channel (t.me/s/<channel>), the page that anyone can open in a browser without a Telegram account. For each channel you give, it saves the newest posts (or the posts that match your search), plus one row about the channel. It can stop at a date, search inside the channel with Telegram's own search, and on a schedule return only posts that are new since the last run. Every channel gets a clear status, so a wrong name is never mixed up with a quiet channel.

Good fit

• Follow what a list of public news, crypto or brand channels posted today.
• Build a dataset of posts with text, views, reactions, media and links.
• Find posts about a keyword or #hashtag inside many channels at once.
• Get subscriber counts and descriptions of many channels.
• Send only new posts to a spreadsheet, a chat or an AI agent every hour.

Not a fit / limits

• Only public channels that have a web preview. Private channels, invite links (t.me/+...), groups, chats, users and bots are not supported.
• No comments, no member or subscriber lists, no DMs, no global Telegram search: only posts of the channels you give.
• Telegram rounds view and reaction counts on the web (for example 830K). The Actor turns them into numbers (830000); it cannot make them more exact than Telegram shows.
• Videos: the web preview usually gives only a preview image, not the video file. Photo links work, but all media links are temporary.
• One run reads about 1,800 posts in full; at about 4.5 minutes it stops and keeps what it has (see Limits).

Quick start

1. Open the Input tab.
2. Add one or more channels to Channels (channels), for example durov and telegram (this is the prefill).
3. Keep Max posts per channel (maxPostsPerChannel) at 20, or set a higher number.
4. Click Start.
5. Open the Output tab: the Posts view shows the post columns (channel, date, text, views, media); the Channels view shows the channel columns (status, title, subscribers). Both views list all rows; the Row type column tells posts and channels apart.

Expected result: 20 posts per channel, newest first, and one channel row each, in about 5 seconds.

Input

Input fields

Field	Plain name	Type	Required	Default / prefill	Allowed values	Effect	Example	Recommendation
channels	Channels	array of strings	yes	prefill ["durov", "telegram"]	1–100 items: name, @name, https://t.me/name or https://t.me/s/name	which channels to read	["bbcrussian", "@telegram"]	Repeats in any spelling are read once. Invalid values are skipped and counted in the run message; the rest still run.
maxPostsPerChannel	Max posts per channel	integer	no	default 20, prefill 20	1–1000	how many posts per channel, newest first	200	Each post row is charged; keep it small for frequent runs.
since	Only posts since	string	no	empty (no date limit)	date 2026-10-01, date-time 2026-10-01T12:00:00Z, or 30 minutes, 24 hours, 7 days, 2 weeks, 1 month	skips older posts and stops reading the channel there	"7 days"	The best way to keep runs short and cheap.
searchQuery	Search inside channels	string	no	empty	up to 200 characters; a word, phrase or #hashtag, any language	returns only posts that Telegram's channel search matches	"bitcoin"	Leave empty to get all posts.
includeChannelInfo	Exact subscriber count	boolean	no	true	true / false	one extra request per channel for the exact number in subscribers	false	Turn off when the rounded subscribersText (10.4M) is enough.
onlyNew	Only new posts since the last run	boolean	no	false	true / false	remembers the newest post per channel and next time returns only newer posts	true	For schedules; see Scheduling and monitoring.
stateKey	Monitoring name	string	no	default	1–40 letters, digits, - or _	keeps separate memory for different schedules	"crypto-news"	Use one name per schedule.

If no value in channels is a valid channel name, the run fails with Invalid input and only the start fee is charged.

Interacting settings

When	Input A	Input B	Effective behavior	Example	Cost/result effect
Both set	maxPostsPerChannel	since	reading stops at whichever comes first	200 and "1 day"	fewer posts and requests
Search and date	searchQuery	since	matches newer than since, newest first	"#news" and "7 days"	fewer posts
Monitoring	onlyNew	maxPostsPerChannel	at most this many new posts per run, oldest new post first; the rest come next run	true and 50	no gaps between runs
Monitoring and search	onlyNew	searchQuery	memory is kept per channel and search; the newest matches come first	true and "ton"	set maxPostsPerChannel above the expected number of new matches
Monitoring and date	onlyNew	since	both apply; posts older than since are skipped	true and "1 day"	—

Input examples

Latest posts of two channels

{
  "channels": ["durov", "telegram"],
  "maxPostsPerChannel": 20
}

Expected behavior: about 4 requests, about 5 seconds. Expected output: 40 post rows and 2 channel rows with subscribers. Cost: $0.0841 on the Free plan.

Last 7 days of several channels, search for a word

{
  "channels": ["bbcrussian", "meduzalive", "nexta_live"],
  "since": "7 days",
  "searchQuery": "Россия",
  "maxPostsPerChannel": 100
}

Expected behavior: only posts that Telegram's search matches, newer than 7 days. Expected output: post rows with searchQuery set; channel rows with status.

Hourly monitoring

{
  "channels": ["durov", "telegram"],
  "maxPostsPerChannel": 50,
  "onlyNew": true,
  "stateKey": "telegram-news",
  "includeChannelInfo": false
}

Expected behavior: the first run returns the newest posts and remembers them; every next run returns only posts published after that. Expected output: new post rows only; with nothing new, the run ends as NO_RESULTS with 2 channel rows.

Output

Results go to the default Dataset. Each row has recordType: post or channel. The dataset has two views, Posts and Channels. A view only picks columns, it does not hide rows: post rows and channel rows are in both views. To keep only posts, filter by recordType = post (for example in Excel, Google Sheets or your code). A run summary is saved as RUN_SUMMARY in the key-value store. The Output tab links all three.

Output example

{
  "recordType": "post",
  "channel": "durov",
  "channelTitle": "Pavel Durov",
  "postId": 552,
  "url": "https://t.me/durov/552",
  "date": "2026-10-11T14:52:40+00:00",
  "edited": false,
  "text": "✨ Designers × AI\n\n🆕 The mockup era is over. …",
  "author": "Pavel Durov",
  "views": 830000,
  "viewsText": "830K",
  "forwardedFrom": "Telegram Designers",
  "forwardedFromUrl": "https://t.me/designers/266",
  "replyToPostId": null,
  "replyToUrl": null,
  "mediaType": "video",
  "mediaUrl": "https://cdn4.telesco.pe/file/…",
  "media": [
    {
      "type": "video",
      "url": null,
      "thumbnailUrl": "https://cdn4.telesco.pe/file/…",
      "durationText": "1:15",
      "title": null
    }
  ],
  "linkPreview": null,
  "poll": null,
  "links": ["https://t.me/designers"],
  "hashtags": [],
  "reactions": [
    { "emoji": "⭐", "paid": true, "count": 5280, "countText": "5.28K" },
    { "emoji": "custom:5936157098181135162", "paid": false, "count": 10700, "countText": "10.7K" }
  ],
  "searchQuery": null,
  "sourceUrl": "https://t.me/s/durov",
  "fetchedAt": "2026-10-14T15:01:48.976Z"
}

This post row is from the first input example; the text, media links and reactions are shortened. null means Telegram shows no value for this post (not a reply, no link preview, no poll). For this video the web preview has only a preview image, so media[0].url is null.

Post fields

Field	Meaning	Type	Can be empty?	Example / enum	Notes
recordType	row type	string	no	post
channel	channel username	string	no	durov	as in the channel link
channelTitle	channel name	string	yes	Pavel Durov
postId	post number in the channel	integer	no	552	with channel, a stable key to remove duplicates
url	link to the post	string	no	https://t.me/durov/552
date	when it was posted	string	yes	2026-10-11T14:52:40+00:00	ISO 8601, UTC
edited	the post was edited	boolean	no	false
text	post text	string	can be ""	✨ Designers × AI…	line breaks kept; "" for media-only posts
author	signature shown under the post	string	yes	Pavel Durov	only if the channel signs posts
views	views as a number	integer	yes	830000	from viewsText; rounded by Telegram
viewsText	views as Telegram shows them	string	yes	830K
forwardedFrom, forwardedFromUrl	original channel or person of a forwarded post	string	yes	Telegram Designers	null if not forwarded
replyToPostId, replyToUrl	the post this one replies to	integer, string	yes	null
mediaType	main media type	string	yes	photo, video, round_video, document, audio, voice, sticker, poll, location, album, unsupported	null for text-only posts
mediaUrl	first media link	string	yes	https://cdn4.telesco.pe/file/…	photo image or video preview image; temporary
media	all media: type, url, thumbnailUrl, durationText, title	array	can be []	see example	url is null when the preview has no file link
linkPreview	link card: url, siteName, title, description, imageUrl	object	yes		null if the post has no link card
poll	question, type, votersText, options[] (text, percent)	object	yes		null if no poll
links	links in the text	array	can be []	["https://t.me/designers"]
hashtags	hashtags	array	can be []	["#news"]
reactions	emoji, paid (Telegram Stars), count, countText	array	can be []	see example	custom emoji appear as custom:<id>
searchQuery	the search that found this post	string	yes	null	null when no search was used
sourceUrl	page the post was read from	string	no	https://t.me/s/durov
fetchedAt	when the run read it	string	no	2026-10-14T15:01:48.976Z	views and reactions are as of this time

Channel fields

One channel row per valid channel, also when the channel is not found. If the run stops in the middle of a channel (time or request limit), that channel gets one more row with status: capped; this row is free.

Field	Meaning	Type	Can be empty?	Example / enum
recordType	row type	string	no	channel
status	result for this channel	string	no	complete, not_found, error, capped
statusCode	why, when not complete	string	yes	NOT_FOUND, NO_PUBLIC_PREVIEW, or an error code
statusReason	plain explanation	string	yes	"telegram" exists on Telegram but has no public channel preview…
channel, url	username and link	string	no	durov, https://t.me/durov
title, description	channel name and description	string	yes	Pavel Durov, Founder of Telegram.
verified	verified badge	boolean	yes	true
avatarUrl	channel photo (temporary link)	string	yes
subscribers	exact subscriber count	integer	yes	10386608 — null when includeChannelInfo is off
subscribersText	subscribers as Telegram shows them	string	yes	10.4M
photos, videos, files, linksShared	media counters from the channel page	integer	yes	102, 47, null, 200
postsReturned	posts saved for this channel	integer	yes
sourceUrl, fetchedAt	page and time	string	no

Run summary (RUN_SUMMARY)

Field	Meaning
status	COMPLETE, PARTIAL (some channels not found or failed), NOT_FOUND (no channel found), NO_RESULTS (channels found, no posts matched), CAPPED (stopped at a limit, rows kept), SOURCE_ERROR, ERROR, INVALID_INPUT
message	the same text as the run status message
requested	the settings used
counts	channels, postRows, channelRows, notFound, errors, chargedPostRows, chargedChannelRows, invalidChannels, unprocessedChannels
channels[]	one entry per channel: channel, outcome (complete, no_posts, not_found, no_public_preview, error, capped, unprocessed), posts, pages, newestPostId, oldestPostId, watermarkBefore, watermarkAfter (monitoring), reason
warnings, http, source, timing	warnings, request counters, data source note, start and finish time

Output entities and surfaces

Situation	Where	Result shape	User action
Normal	Dataset, Posts and Channels views	post rows + one channel row per channel	open, export or call the API
Channel does not exist	Channels view, RUN_SUMMARY	status: not_found, statusCode: NOT_FOUND	check the name
Group, user, bot or preview turned off	Channels view	not_found / NO_PUBLIC_PREVIEW	only public channels work
No posts in the period or no search match	RUN_SUMMARY	NO_RESULTS, channel rows complete, outcome no_posts	widen since or change the search
Stopped at a limit	Dataset + RUN_SUMMARY	rows kept, CAPPED, cut channel status: capped, unread channels unprocessed	split the list into smaller runs
Telegram not answering	RUN_SUMMARY, run FAILED	SOURCE_ERROR (rows kept) or ERROR	try again later

Runs with COMPLETE, PARTIAL, NOT_FOUND, NO_RESULTS or CAPPED end as SUCCEEDED; SOURCE_ERROR, ERROR and INVALID_INPUT end as FAILED.

Pricing

Pricing unit

You pay per event. Prices per 1,000 events:

Event	Free	Bronze	Silver	Gold	Platinum	Diamond
Post — one saved post row	$2.00	$1.60	$1.30	$0.40	$0.40	$0.40
Channel — one channel row (found or not found)	$2.00	$1.80	$1.50	$1.00	$1.00	$1.00

Each run also has a start fee of $0.0001. Platform usage is included.

• A channel that Telegram answered about is charged once: found (complete) or not found (not_found). A channel row with status: error is free.
• Invalid input values are not charged. A run with no valid channel pays only the start fee.
• A run with no matching posts, or a monitoring run with nothing new, pays only for the channel rows and the start fee.
• If you set a maximum cost per run, the Actor stops before it goes over it and keeps the rows already saved.

Cost examples (Free plan prices)

Scenario	Input scope	Events	Charge	Time
Prefill	2 channels × 20 posts	2 channels + 40 posts	$0.0841	about 5 s
Medium	5 channels × 100 posts	5 + 500	$1.0101	about 1 minute
Large	10 channels × 180 posts	10 + 1,800	$3.6201	about 3.5 minutes
Nothing new	2 channels, onlyNew, no new posts	2 channels	$0.0041	a few seconds
Wrong names	3 channels not found	3 channels	$0.0061	a few seconds

On the Gold, Platinum and Diamond plans the prefill run costs $0.0181.

Use cases

Buyer job	Input path	Output entity	Key fields	Next integration
Hourly news or crypto monitoring	channels + onlyNew	post	date, text, url, views	Schedule + webhook, Google Sheets, Slack
Post engagement research	channels + maxPostsPerChannel	post	views, reactions, forwardedFrom	CSV / Excel export
Keyword tracking	channels + searchQuery + since	post	text, date, searchQuery	API, spreadsheet
Channel directory	channels + includeChannelInfo	channel	subscribers, title, description, verified	API, spreadsheet
AI agent context	channels (small)	post	text, date, url	MCP

Scheduling and monitoring

Create a schedule (for example every hour) with onlyNew: true and a stateKey such as telegram-news. The first run returns the newest posts and remembers the newest post of each channel. Every next run returns only posts published after it, oldest first, without gaps: if more posts arrived than maxPostsPerChannel, the rest come in the next run. The memory is kept in your account, in the key-value store telegram-channel-scraper-state; use a different stateKey for each schedule. A run with nothing new ends as NO_RESULTS and costs only the channel rows. Add a webhook or an integration to send new rows to Slack, a sheet or your own API when the run finishes.

Without onlyNew, use since: "1 day" and remove duplicates by channel + postId.

Switching from another Telegram scraper

Some Telegram scrapers return rows like username, fullName, followers and a message object, and give views as text. This Actor uses flat fields and numbers:

Other scraper	This Actor
username	channel
fullName	channelTitle (post row), title (channel row)
followers	subscribers (exact) and subscribersText
message.fulldate	date (ISO 8601, UTC)
message.views (1.16M as text)	views (1160000) and viewsText
message.description	text
message.link	url
preview	linkPreview
image, video	media[] with type, url, thumbnailUrl

Channels that are not found get their own row with status: not_found and a reason, not an empty row.

API, export and integrations

• API: start a run with POST https://api.apify.com/v2/acts/cliqtomedia~telegram-channel-scraper/runs and the input JSON; read rows from the run's Dataset and keep the rows with the recordType you need (?view=posts or ?view=channels only picks the columns), and read the summary from the RUN_SUMMARY record.
• Export: JSON, CSV, Excel, XML or RSS from the Dataset.
• Webhooks / no-code: Apify integrations (Make, Zapier, Google Sheets, Slack) can send new rows when a run finishes.

Use with AI agents (MCP)

• MCP server with only this Actor: https://mcp.apify.com?tools=cliqtomedia/telegram-channel-scraper.
• Smallest useful input: {"channels": ["durov", "telegram"], "maxPostsPerChannel": 20}.
• Key fields: text (what was posted), date (when), url (link to cite), channel (where), views (reach as a number).
• Two channels with 20 posts take about 5 seconds and cost about $0.08 on the Free plan. One run reads up to about 1,800 posts; use since or searchQuery to keep answers short.

Limits and partial results

• Up to 100 channels per run and 1,000 posts per channel. Telegram's web preview gives about 20 posts per page, and the Actor reads about one page per second.
• Each run stops reading Telegram after about 4.5 minutes, whatever the run timeout is. That is enough for about 1,800 posts. If the time runs out, saved rows stay, the run is CAPPED, the channel being read is capped and the rest are unprocessed. Split large lists into several runs.
• If you set a maximum cost per run, the Actor stops before going over it: rows already saved stay, the run is SUCCEEDED with CAPPED in RUN_SUMMARY.
• View, reaction and subscriber texts are rounded by Telegram; only subscribers with includeChannelInfo is exact.
• Media links (telesco.pe) expire after some time. Download the files soon if you need them.

Troubleshooting

A channel returns no posts

Check its channel row (recordType = channel) in the Channels view: not_found / NOT_FOUND (wrong name), not_found / NO_PUBLIC_PREVIEW (a group, user, bot or a channel with the web preview turned off), or complete with outcome no_posts in RUN_SUMMARY (nothing matched since or searchQuery).

The run ends as CAPPED

Read RUN_SUMMARY.message: either the maximum cost you set was reached, or the run reached its 4.5-minute reading limit. Saved rows stay in the Dataset. Lower maxPostsPerChannel, use since, or split the channel list.

Some of my values were skipped as invalid

Only channel names (3–32 letters, digits or _, starting with a letter), @name and t.me links work. Invite links (t.me/+...) and other websites are skipped; the run message says how many.

The run fails with SOURCE_ERROR or ERROR

Telegram did not answer or asked the Actor to slow down several times. Rows saved before the error stay in the Dataset. Try again later.

FAQ

Do I need a Telegram account, phone number or API key?

No. The Actor reads only the public web preview that Telegram shows to anyone, without login.

Why are views like 830000 and not exact?

Telegram shows rounded counts on the web (830K). The Actor gives you the number (views) and the original text (viewsText). Exact view counts are not public.

Can I get comments, members or posts from private channels?

No. These are not in the public web preview.

Are the counts live?

Views, reactions and subscribers are as they were at fetchedAt. Run again to refresh them.

Is scraping Telegram allowed?

The Actor reads only public pages that Telegram shows without login, at a slow rate, and does not bypass any login or CAPTCHA. You are responsible for how you collect, store and reuse the content, including personal data and authors' rights, under Telegram's terms and the laws that apply to you.

Learn more

• Telegram channel web preview example: the page the Actor reads.
• Telegram Terms of Service.
• Apify pay-per-event pricing.
• Apify schedules.

Support

Use the Issues tab of this Actor for bugs, questions and feature requests. Include the run ID, the input and the error text. Never post tokens, cookies or private data.