# Overview (/docs/engines) Engines are runs that check, keep only what is new and deliver it, and each one can price its request for free before it runs Engines are runs that check, keep only what is new and deliver it. The most common thing people do by hand is repeat the same request on several days. A Monitor does that for you, so it comes first. Every engine below except Ask AI can price a request for 0 credits before it runs, and Ask AI quotes before it spends. Nothing starts until you send the real call. | Engine | What it does | What it costs | | --- | --- | --- | | [Monitors](/docs/engines/monitors.md) | Repeats one call on a schedule and keeps the numbers you choose | The call's own price plus 1 credit a run | | [Listening](/docs/engines/listening.md) | Watches keywords, a topic or accounts and delivers each new mention once | The searches each check makes, up to a monthly cap you set | | [Search everywhere](/docs/engines/search-everywhere.md) | Sends one query to many platforms and returns one ranked list | 20 credits a search | | [Discover](/docs/engines/discover.md) | Searches many platforms inside a date window and scores each row against your question | Each platform search at its own price, plus scoring | | [Find creators](/docs/engines/find-creators.md) | Finds and ranks the creators who post about a topic | Searches, lookups and scoring, never past your `max_credits` | | [Cohorts](/docs/engines/cohorts.md) | Reads what a list of accounts you chose posted about your keywords | Free to create and upload, then a price for each page a query reads | | [Ask AI](/docs/engines/ask-ai.md) | Does any of this from plain words, with no code | The credits of the calls it makes, quoted first | ## Which one should you start with? Start with a Monitor if you know the account or post you care about. Start with Listening if you care about a topic or a brand. Use Search everywhere for a quick look across platforms. Use Discover when you need a date window and a verdict on every post. Use Find creators when you need people, not posts. Use Cohorts when you already have the list of people. Use Ask AI when you would rather not write code. ## How does a dry run work? Add `dry_run=1` to the request, or `dry_run: true` in a JSON body. The answer holds the plan and the price. It costs 0 credits and creates nothing. The page of each engine opens on one. ## The engines - [Monitors](/docs/engines/monitors.md): Repeat a call on a schedule and keep only what changed. - [Listening](/docs/engines/listening.md): Get each new mention of a topic once. - [Search everywhere](/docs/engines/search-everywhere.md): One query, many platforms, one ranked list. - [Discover](/docs/engines/discover.md): A date window and a score on every post. - [Find creators](/docs/engines/find-creators.md): A ranked list of the people behind a topic. - [Cohorts](/docs/engines/cohorts.md): Which of your accounts talked about it. - [Ask AI](/docs/engines/ask-ai.md): Ask in plain words and approve the price.