Apify announced today, November 10, 2026, two things that matter if you build AI agents: its AI assistant can now build whole workflows from a sentence, and its Actors — the little web tools agents call — can now stay permanently switched on. The second one is the bigger deal. Server Actors are generally available, which means Apify's marketplace of over 100,000 Actors just stopped being a library of jobs you start and start again, and started being infrastructure that answers instantly.
I haven't run Server Actors myself. Nobody outside Apify has had them in general availability before today. This article is built from Apify's announcement, released at Run, the company's first user conference, and checked 10 November 2026. Where a number is Apify's, I'll say so — because a launch-day claim is a vendor claim, not an independent result.
What Server Actors actually change
Here's the distinction that matters. A normal Actor is a job: your agent asks for something, Apify spins up an execution, the job runs, you get the result. That works fine for scraping a site or processing a big dataset. It doesn't work when your agent needs an answer right now — a tool call mid-conversation, a real-time lookup, a check that has to happen a hundred times a minute.
Server Actors are the always-on version. They deploy as services on the Apify platform and behave like APIs: your application or agent sends a request and gets a response, without starting a new execution each time. No cold start, no job queue, no waiting for a run to spin up.
That sounds like a small infrastructure detail, and in a way it is. But it's the detail that decides whether an agent feels instant or feels like it's filing paperwork. Apify's bet is that agents moving from demos into production need both modes in one place — the heavy batch jobs and the always-ready services — and Server Actors are the second half of that promise. They're available now in Apify Store.
Apify AI: from finding tools to building them
The other half of the announcement is Apify AI, the conversational interface inside Apify Console that first appeared in July 2026. Back then it was a way to discover and run Actors by chatting instead of browsing a catalogue. Now it goes further: you can execute code, orchestrate multiple Actors, and build entirely new Actors using natural-language instructions, without manually connecting tools or transforming outputs.
The part I'd actually pay attention to is what happens to the workflow afterwards. Unlike a one-off AI answer, the thing you build becomes a reusable Actor — something you can schedule, plug into other applications, or publish to Apify Store. Apify AI also got broader access to MCP tools, including tasks and schedules, so the workflows it builds can reach further into automation.
Put plainly: Apify AI used to help you find a tool. Now it helps you build the tool, wire it into a workflow, and keep it running. That's a real step up from a search box with a chat window — if it works as described, which nobody outside Apify has verified yet.
The numbers behind the launch
Apify's growth figures explain why this launch is happening now. Weekly Actor runs passed 267 million as of October 2026 — up more than fivefold from 50 million a week at the end of 2025. In the same month, Apify Store passed 100,000 Actors, built by a community of more than 4,500 independent developers and official partners. Those creators collectively earn more than $1.7 million a month through the platform.
There's also a distribution number worth noting. Apify's official plugin in Claude's directory ranks among the 50 most-used plugins as of October 31, 2026, with more than 57,000 lifetime installs and over 12,000 active accounts. And Apify connects to any AI agent via mcp.apify.com, so this isn't a Claude-only story.
Those are Apify's numbers, from its own announcement. The direction is hard to argue with — fivefold growth in weekly runs in under a year is not a rounding error — but the exact figures are vendor-reported, and the "largest marketplace of trusted tools for AI" line is Apify describing itself. Take the trend seriously; take the superlative with a pinch of salt.
Who this is actually for
If you're wiring agents into real workflows, this is a practical launch. The use cases Apify names — competitive monitoring, lead generation, recruitment research, software development — are the ones where an agent needs live web data right now, not a batch job that finishes in ten minutes.
The most interesting reader for Server Actors is the one who already has an Apify workflow in production and has been working around the cold-start problem: polling a job, caching results, building a wrapper service in front of an Actor. If that's you, a Server Actor is the thing you were building yourself, now offered as a platform feature. The test is simple: point your agent at one and see whether the latency you were fighting disappears.
For everyone else, the honest answer is to watch. Apify AI building Actors from natural language sounds great in an announcement; the question is what it produces when the workflow is messy, the website changes its layout, and the data comes back in a shape nobody expected. That's the test that decides whether this is a real capability or a conference demo.
The catch, stated plainly
- No pricing in the announcement. Apify says nothing about what Server Actors cost to run, or how Apify AI usage is billed. An always-on service has an always-on price, and until that's published, you can't do the math on whether it beats running your own wrapper.
- Every number is Apify's. The 267 million weekly runs, the 100,000 Actors, the $1.7 million a month to creators, the Claude plugin ranking — all from Apify's own release. The growth trend looks real, but none of it is independently verified.
- Apify AI's workflow-building is untested outside Apify. "Build entirely new Actors using natural-language instructions" is a strong claim. The first time a generated Actor breaks on a real website, you'll find out how good the error handling is. I haven't seen that yet, and neither has anyone else.
- Always-on is a new failure mode too. A job that runs and finishes is easy to reason about. A service that stays up needs monitoring, and when it goes down, everything that depends on it goes down with it. Server Actors move that responsibility onto Apify's platform — which is the point, but also the risk.
The bottom line
Server Actors are the least flashy kind of launch and possibly the most useful one Apify has made: the same tools, now available the instant an agent asks for them. Combined with Apify AI building workflows from plain language, the direction is clear — Apify wants to be the place where agents find tools, build tools, and run them, all in one platform.
The numbers say the demand is there. Whether Server Actors deliver on the latency promise is a test you can run this week, and it's a cheap one: pick the Actor your agent already calls most often, deploy it as a Server Actor, and measure what changes.
Have you hit the cold-start wall with agent tools — or is the job model still fast enough for what you're building?
Sources: Apify via GlobeNewswire, "Apify Unveils New AI Agent Capabilities and Server Actors to Support Growing Demand" (https://www.manilatimes.net/2026/11/10/tmt-newswire/globenewswire/apify-unveils-new-ai-agent-capabilities-and-server-actors-to-support-growing-demand/2466557), 10 November 2026. All capability, growth and earnings claims are Apify's, stated as such. No hands-on testing was performed for this article. Nothing here is sponsored — no affiliate links on this page. This breaking article is not the daily buyer-intent post. Plain link for readers: apify.com.
Top comments (0)