What Reddit search never did
Reddit's search box matches titles and a little body text, ranks by its own relevance, and leaves comments out. For anyone watching a brand, a ticker, a competitor or a phrase, comments are where most of the mentions are. Redlisten indexes the full text of every public post and comment and searches it with plain operators, newest at the top.
Paid plans search the last 30 days and free keys search the last 7 days. Both windows roll forward every day.
The query syntax
Everything goes in q. Words are matched as whole words, case does not matter, and terms such as c++, c# or at&t match as typed.
- Space or AND between words means all of them must appear:
notion alternativeandnotion AND alternativeare the same query. - OR means any of them:
hubspot OR pipedrive OR "zoho crm". - NOT or a leading minus drops matches:
crm NOT salesforce,crm -salesforce. - Quotes make a phrase:
"road closed"matches those two words together, in that order. - subreddit: limits to a community, and can be negated:
subreddit:smallbusiness,NOT subreddit:memes. - author: limits to one account:
author:example_user. - Brackets group:
(hubspot OR pipedrive) "too expensive".
Alongside q, three parameters narrow the result set. type is link for posts or comment for comments, and both when omitted. t is hour, day, week or month, with month meaning your plan's window. limit and after page through results, up to 100 a page as on Reddit, or 50 on a free key.
# Posts and comments mentioning a product, outside two communities, this week
curl -G "https://api.redlisten.com/search" \
-H "Authorization: Bearer rl_live_YOUR_KEY" \
--data-urlencode 'q="notion alternative" OR "notion replacement" NOT subreddit:Notion NOT subreddit:productivity' \
--data-urlencode 't=week' \
--data-urlencode 'limit=100'
# Comments by one author that mention a ticker
curl -G "https://api.redlisten.com/search" \
-H "Authorization: Bearer rl_live_YOUR_KEY" \
--data-urlencode 'q=$NVDA author:example_user' \
--data-urlencode 'type=comment'
The response is a Reddit style Listing, so the same parser that reads /r/{sub}/new.json reads search results. The path also works as /r/{sub}/search?q= to scope a query to one community, and as /search.rss?q=&key= for a feed reader. See the docs for the field list.
Searching comments
Set type=comment and the same query runs over comment bodies. Each result carries link_id and parent_id, so a hit can be expanded into its thread with one call to /comments/{id}. That is the shape most monitoring tools want: the matching comment, the post it sits under, and the replies around it.
A search call counts as one request however many results it returns. There are no daily caps; the monthly request quota is the meter. A Pro key with 50,000 requests a month runs a query a minute, all month, with room left.
Turning a query into a stream
A search answers the question once. A keyword stream keeps answering it. Create a stream with the same words, split into any, all and exclude lists, plus the same community and author filters, and Redlisten collects every new post and comment that matches from then on. On paid plans the stream fills with the last 30 days of matches when it is created, so the history and the live tail are in one place.
Read a stream's matches with a cursor on your own schedule, or have each match delivered: a webhook POST per match, a message in a Slack or Discord channel, or a daily or weekly email digest. Free keys get one stream with email digests; Pro keys get ten streams with every delivery type. The keyword alerts page walks through the setup, and the docs list every field.
Counting mentions over time
For a report rather than an alert, /v1/trends takes a query and returns a count of posts and comments per hour or per day across the window. Team keys can compare several terms in one call and list the communities where a term appears most. That turns "how is the launch landing" into a chart with one request.