Telegram Keyword Search Scraper

Telegram Keyword Search and Channel Scraper

Turn public Telegram content into a dataset you can search, review and export. Discover Telegram pages by keyword, or collect recent posts from public channels you already know — without writing code or supplying Telegram API credentials.

Use the results for brand research, content planning, news discovery and channel comparisons. Available data includes post text, dates, views, reactions, media links and channel metadata, depending on what Telegram exposes in its public web preview.

Ready to start? Choose a ready-made example below, customize the keywords or channel usernames, and click Start.

Choose your starting point

Ready-made task	What you can do
Telegram AI News Search	Discover indexed Telegram pages about artificial intelligence.
Telegram Brand Mentions Search	Start with ChatGPT, then replace it with your brand or product.
Telegram Remote Jobs Search	Find job-related Telegram content for manual review.
Telegram Channel Posts Export	Export recent posts from a public channel by username.
Telegram Multiple Channels Comparison	Collect posts from several channels for comparison in a spreadsheet.

These are editable starter configurations, not separate products. Search results change over time. The Actor collects data; it does not verify job offers, calculate sentiment or generate a comparison report.

How to scrape Telegram without code

1. Choose Keyword search to discover content, or Channel to read known public channels.
2. Enter one or more keywords, or channel usernames such as telegram. An initial @ is optional; use usernames rather than full URLs.
3. Start small: use 10 search results per keyword or 20 messages per channel.
4. Click Start, then open the run's dataset to review the results.
5. Export your dataset as JSON, CSV or Excel using Apify's export options. JSON preserves nested objects such as reactions and channel metadata.

Save your input as an Apify task to reuse it. You can add an Apify schedule for recurring collection; the example tasks do not enable a schedule automatically.

Keyword search or channel scraping?

Your goal	Mode	How it works
Discover Telegram content about a topic, brand or product	keyword	Searches Google's index for Telegram URLs, then extracts a post from each accessible public preview page.
Collect recent posts from known public channels	channel	Reads channel web previews and paginates backwards until the message cap, date cutoff or available history is reached.

What keyword search returns

Keyword mode is a discovery tool for indexed Telegram pages, not an exhaustive search of all Telegram messages. The Actor extracts the newest message on each discovered preview page. If a search result points to a channel homepage, the extracted post may differ from Google's matching snippet and may not contain your keyword.

Search date, country and language settings influence discovery. They do not guarantee the date, location or language of every extracted post. Review the returned text and dates when exact matching matters. The same post can appear from different URLs, keywords or repeated runs.

What channel mode returns

Channel mode collects the most recent accessible messages up to your cap, then returns the collected messages oldest first within each channel. It can paginate through available public history, but cannot guarantee a complete archive. Private channels, private groups, member lists and content unavailable in the public web preview are not supported.

Input settings

Keyword search settings

Field	What to enter
mode	Set to keyword.
keywords	One or more search phrases. Each entry starts a separate search.
maxResultsPerKeyword	10–100; default 100. Caps search-result URLs examined per keyword, not guaranteed output rows.
afterDate	Optional search lower date bound, inclusive. Use YYYY-MM-DD or a relative value such as 30 days.
beforeDate	Optional search upper date bound, exclusive. Use YYYY-MM-DD.
countryCode	Search-region preference; default us. It does not filter channels by their owners' location.
languageCode	Search-language preference; default en. It does not translate posts.

Channel settings

Field	What to enter
mode	Set to channel. This is the input form's default mode.
channels	Public channel usernames, for example telegram and durov.
channelMaxMessages	Maximum messages per channel: 10–100,000; default 100. Actual output can be lower.
channelAfterDate	Optional inclusive cutoff: YYYY-MM-DD or a relative value such as 30 days. Dated posts before this cutoff are excluded.

Relative dates support days, weeks, months and years. Months use 30 days and years use 365 days. Keep the mode explicit in API requests and use the settings for your chosen mode.

Example inputs

You can enter these settings in the visual form; JSON is only needed if you prefer the JSON editor or API.

Discover AI-related Telegram content

{
  "mode": "keyword",
  "keywords": ["artificial intelligence"],
  "maxResultsPerKeyword": 10,
  "countryCode": "us",
  "languageCode": "en"
}

Export recent posts from a public channel

{
  "mode": "channel",
  "channels": ["telegram"],
  "channelMaxMessages": 20
}

To collect a rolling date window, add "channelAfterDate": "30 days". Increase the message cap if a busy channel publishes more posts than your current limit during that window.

Output: what is in your Telegram dataset?

Field	Meaning
id	Message ID within its channel; not globally unique.
text, date	Available post text and publication timestamp.
source_name	Channel display name.
source_id, sender	Numeric channel identifier, when exposed; sender is represented as a string.
view_count	Visible view count; abbreviated counts are converted to numbers and may be approximate.
reactions	Available emoji reaction counts.
media_url	One available media URL; not a complete album or media download.
pinned	Whether the preview identifies the post as pinned.
channel_info	Available subscriber/media/link counters and channel description.
source_url	Preview URL in keyword mode; omitted in channel mode. It may refer to a preview page rather than a direct post link.
processed_at	Time the record was extracted.
type, source_type, silent	Compatibility fields currently emitted as Regular, Channel and false; silent is not a verified notification setting.

Missing information can appear as null, an empty string or an empty object. Channel counters vary by page. In keyword mode, missing post text or dates may fall back to Google's snippet or date.

Example output record

Illustrative keyword-mode record — values below are examples, not measured channel statistics:

{
  "id": 123,
  "type": "Regular",
  "date": "2026-09-01T10:00:00+00:00",
  "text": "An example public channel announcement.",
  "sender": "1234567890",
  "silent": false,
  "pinned": false,
  "view_count": 2500,
  "reactions": { "👍": 42 },
  "media_url": "",
  "source_id": 1234567890,
  "source_name": "Example Channel",
  "source_type": "Channel",
  "source_url": "https://t.me/s/example_channel?before=124",
  "channel_info": {
    "subscribers": 12000,
    "photos": 150,
    "videos": 20,
    "files": 3,
    "links": 85,
    "description": "Example channel description."
  },
  "processed_at": "2026-09-01T10:05:00.000Z"
}

For repeated collection, deduplicate downstream using the channel identifier and message ID when available. A message ID alone is insufficient across multiple channels.

Practical use cases

• Brand research: discover indexed pages related to your brand, review the actual post text, and collect relevant examples for your team.
• News and topic discovery: find public channels discussing your subject, then use channel mode to collect their accessible recent posts.
• Content planning: examine publication dates, text and visible engagement to identify topics worth investigating.
• Channel reporting: export several channels to a spreadsheet and calculate your own activity or engagement summaries.
• Research datasets: collect accessible public posts for downstream classification, summarization or trend analysis in your own tools.

Costs and sensible starting limits

Check the Actor's current pricing on its Apify Store page and the usage recorded for your runs. No Telegram API credentials are required, but running on Apify is subject to the applicable Actor and platform charges.

Keyword mode also launches apify/google-search-scraper as a separate Actor run. Its charges can add to the parent run's cost. Review both runs; do not assume a spending limit on the parent covers the child search run.

Start with one keyword and 10 results, or one channel and 20 messages. More keywords and larger channel histories increase work and may increase cost. Output counts and completion times depend on search coverage, channel availability and network conditions.

Frequently asked questions

Can I search every Telegram message?

No. Keyword mode discovers Telegram URLs available through Google's index. Channel mode reads accessible public web previews. Neither mode can promise complete Telegram coverage.

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

No Telegram credentials are requested. You run the Actor through your Apify account.

Why did a successful run return fewer results or no results?

A successful status does not guarantee records. Search results may be sparse, channels may lack accessible previews, filters may be too narrow, or individual fetches may fail. Check the dataset and run logs. Try fewer restrictions or a known public channel such as telegram.

Can I monitor a channel automatically?

Yes: save the input as a task and configure an Apify schedule. Each run creates its own dataset. Combine and deduplicate results downstream; automatic alerts and cross-run deduplication are not built into this Actor.

Does it download every photo, video or attachment?

No. It returns an available media link when exposed by the preview. Media links can change or expire; complete album downloads are not included.

Can I connect the output to another tool?

Use Apify dataset exports, the Apify API or supported integrations to pass results to spreadsheets and automation workflows. Nested reaction and channel data are easiest to preserve in JSON.

Support and more scrapers

For a reproducible issue, open the Actor's Issues tab with the run ID, non-sensitive input and expected result. This helps distinguish search-coverage limitations from extraction problems.

Build your next research workflow: Discover more Actors by lofomachines for additional data collection and enrichment tools.