To search Telegram posts by keyword, enter a term, choose a date range, and check that the term appears in the post text. This walkthrough takes a search for ChatGPT across public Telegram channels and exports the matching posts to CSV, with a link back to each source.
I built the Apify Actor used here. You’ll need an Apify account to run the example. You can search the public sources without signing in to Telegram. Open the prefilled Telegram keyword search
1. Search public Telegram channels by keyword
Open the prefilled search example. It looks for ChatGPT in public posts from the last 30 days and returns up to 10 posts, with up to 2 per channel. Start with that sample, or swap in a keyword from your own research.
If you start from the Actor’s JSON input editor, this configuration explicitly requires the match to appear in the post:
{
"collectionMode": "keyword",
"keywords": ["ChatGPT"],
"channels": [],
"afterDate": "30 days",
"maxResults": 10,
"maxResultsPerChannel": 2,
"discoveryCoverage": "standard",
"keywordMatchScope": "postText",
"includeContextPosts": false
}
Leave the channel list empty to discover public sources for your keyword. Before clicking Start, set Maximum cost per run in the run options. The post limit controls the number of rows saved, while the spending cap limits the run charge. Actor input reference
2. Verify that the keyword appears in the post
A channel with “ChatGPT” in its name can publish posts about unrelated tools. Post text only requires a literal match in the post text or caption. In Output, inspect a few rows before widening the search.
| Field | What it tells you |
|---|---|
matchedTextKeywords |
Which terms appeared in the post text or caption |
text |
The context around the match |
publishedAt |
When the post was published |
channelUsername |
Which public channel published it |
messageUrl |
Where to open the original post |
If the keyword is present but the posts are too broad, try a more specific phrase, such as ChatGPT pricing. Keep the smaller sample until the results match the question you’re trying to answer.
3. Export Telegram search results to CSV
On the completed run, open Output → Export and select CSV. In Sheets or Excel, sort by publishedAt to review the newest matches. You can also group by channelUsername to see which sources appear in your sample. Keep messageUrl in the spreadsheet so you can trace each row back to the original post.
What the ChatGPT search returned
My November 19, 2026 run returned 10 public posts. The Output tab below shows the first three, along with their publication dates, channels and original links.
Here are those three rows from the CSV export. I kept five columns so the example fits on screen. Apify writes the first value of the matchedTextKeywords array under matchedTextKeywords/0 in this export.
publishedAt,channelUsername,matchedTextKeywords/0,views,messageUrl
2026-11-19T09:18:03.000Z,ai_news_cn,ChatGPT,159,https://t.me/ai_news_cn/45944
2026-11-19T08:08:01.000Z,ai_news_cn,ChatGPT,231,https://t.me/ai_news_cn/45941
2026-11-02T09:58:06.000Z,chatgptme,ChatGPT,3.39K,https://t.me/chatgptme/1977
What if the search returns few posts?
Widen the date range or try another term the authors might use. If you already know the relevant channels, choose Search within known channels and supply both the keyword and the channel usernames. Keyword discovery covers a sample of accessible public posts. Use the export to review the matches you found, and treat missing results as inconclusive.
If you want notifications on future matches, there’s a separate walkthrough for that workflow. Set up Telegram keyword alerts



Top comments (0)