You know what you want. You just don’t know what it’s called.
That’s the gap most people hit the first time they open Apify Store. You’re looking for a tool that does what you need: track a competitor’s prices, build a list of every restaurant in a city with its rating, or watch a subreddit for new posts. You’re not looking for an Actor by name, but that’s exactly what a keyword search needs. So you type what you’d say out loud, it doesn’t quite land, and you end up guessing at search terms.
Say you do find something promising. You open it and hit a second wall: an input form with 20 fields, half of which mean nothing to you. That’s where a lot of people give up.
So two translations stand between you and a result: your problem into Store keywords, then your intent into a configuration form. Both force you to think in the platform’s terms instead of your own. Apify AI removes both steps: describe what you want in plain English, and it finds the right Actor, fills the input form, runs it, and hands back the data, all in one chat.
What Apify AI does #
Apify AI is a conversational assistant built into Apify Console. There’s no separate app, no client to connect, and nothing to install. It’s there the moment you log in.
You’ll meet it in two places at launch:
Apify Store search bar. Type a long, intent-heavy query like “I need to collect guest reviews for a restaurant chain,” and it routes you toApify AI. Type “Instagram” and you get a normal keyword search. The point is that search now reads what youmean, not just the words you typed.Dashboard widget, so you can start a request without going to the Store first.
A dedicated sidebar tab will follow as more capabilities ship.
In beta, Apify AI can already find Actors that match your goal, compare a few of them, pull up an Actor’s details, run it, and return the dataset. That’s the whole loop: discover, decide, run, get data, without leaving the chat.
How it works, and how it differs from Apify’s MCP #
Here's a typical run. You type a prompt, say, "Track my competitor's prices on Amazon." Apify AI picks a suitable Actor, fills in the input schema for you, starts the run, and returns the results with direct links back to Console.
That input schema is where users typically stall. With Apify AI, you never see the 20-field form - you describe the outcome, and the configuration happens in the background.
If you’ve used Apify’s MCP server, this will feel familiar. The Model Context Protocol (MCP) is the interface that lets AI assistants like Claude and ChatGPT find and call Actors directly. Apify AI runs on the same tools underneath. The real difference is the interface: the chat is built into Console, so there’s no setup and no client to wire up, plus you get Apify-rendered UI - live run progress, result previews, and links straight into your account. One honest caveat: if you’re building something programmatic - an automation, an agent, anything that lives outside Console - the MCP server or the CLI is still the right tool. The chat is the zero-setup, in-Console option for getting a job done. MCP is for wiring Actors into your own systems.
For Actor creators: how do I show up in Apify AI? #
There’s no trick to it. When someone makes a request, the agent turns it into a query and ranks Actors much like search does.
Getting suggested comes down to three things:
- A strong Actor quality score
- A clear README
- Well-documented input and output schemas
Get those right, and your Actor shows up more often - in search, in Apify AI, and through MCP. A clear, well-documented Actor is a better match for what someone’s trying to do.
See it as a new activation channel: because the AI fills in messy input schemas and gets people to a finished run, more first runs actually complete. That matters most if you’ve published a high-value Actor with a complex input that users tend to bounce off.
Two kinds of Actors are intentionally left out:
Full-permission Actors are excluded for security. These can reach a user’s entire account and data, so Apify won’t let the model pick and run one on someone’s behalf. That’s a decision a user should make and own personally.Rental Actors are out too, because they don’t fit the sporadic, on-demand nature of these runs.
If you want your Actor eligible for Apify AI, it’s a good moment to move away from full-permission and rental setups.
Search got personal, too #
Console search is now personalized, closer to a recommender system than a classic keyword lookup. It learns from what you and others actually run, so the most useful Actors rank higher, even when they don’t match your exact words.
What’s next #
This is still in beta, and for now, it’s focused on the first job: finding the right Actor and running it. The next iterations will bring tasks, schedules, publishing Actors, and external integrations into the same chat. Today, Apify AI finds and runs. Soon, it’ll handle the entire workflow from a single conversation. Take a look at the Apify AI docs to learn more.
If you’ve ever given up on a search because you couldn’t find the right words, or closed an Actor because the input form was too much, this is built for you. Open Apify Console, type what you’re trying to do, and see what comes back. Because it’s early, your feedback helps shape what comes next: give any response in the chat a thumbs-up or thumbs-down to tell us what worked and what didn’t.