Research tools
Research a topic
Search Reddit, collect matching discussions and rank customer language in one call.
| Tool name | Cost | Plan |
|---|---|---|
research_topic | 1 credit per 5 threads | Every plan |
Use customer wording for the keyword and narrow the subreddits when the topic is broad. Add a brand to save the research to its workspace. From October 30, 2026, paid plans pay 1 credit per 5 threads delivered, so the default 10 threads cost 2; until then a run costs 1.
Inputs
keyword(string, required, up to 300 characters): Topic to research.brand(string, optional, up to 200 characters): Optional brand (name or id) to attach this scrape to. The thread then feeds that brand's research runs directly. Errors if no matching brand exists.limit(number, optional, default10, 1 to 15): Max threads to search and scrape. Default 10, max 15 (higher values are clamped). From October 30, 2026 every 5 threads delivered cost 1 credit on paid plans.subreddits(array of strings, optional, up to 5 items): Optional. Restrict the research to these subreddits (names without r/, e.g. ["running","xxfitness"]). Omit to search all of Reddit. Max 5.exclude_scraped(boolean, optional, defaulttrue): When true (default), threads already in this account's history are not re-scraped; they return in alreadyScraped with scrapeIds for get_history_item. Set false to re-scrape known threads (e.g. to catch new comments on them).mode(string, optional, defaultresearch): Payload size mode. Default research returns a compact subset. Use full only when you need every scraped comment. Values:research,full.max_comments(number, optional, default150, 1 to 2000): Upper bound on comments returned across all scraped threads. Research mode is additionally capped by a serialized-size budget so the result always fits one response, and returns the ranked, thread-balanced top subset (long comment bodies are trimmed with truncated:true; the permalink keeps the full text one click away). Full mode returns up to 2000 (the hard max). The response echoes totalCommentsScraped vs commentsReturned so truncation is always visible.min_score(number, optional, default2, -10000 to 10000): Minimum comment score to return. Default 2 in research mode. Omitted in full mode means no score filter.max_depth(number, optional, default5, 0 to 100): Maximum Reddit reply depth to return. Default 5 in research mode. Omitted in full mode means no depth filter.sort(string, optional, defaulttop): How to order returned comments. Default top. Values:top,new,old,controversial.include_body_only(boolean, optional, defaulttrue): When true, return compact comments with text, score, depth, permalink, and thread context only. Default true in research mode.
Cost and plan
Cost: 1 credit per 5 threads. Plan: Every plan. Cost and plan are separate: a tool can cost no credits and still need a paid plan. See credits and limits.
Use it in chat
You do not need to type the tool name. Connect Adlicio to Claude or to ChatGPT, then ask for the job in plain words. Your AI app picks research_topic when it fits the request.