Apify

In April 2026 it began retiring fixed monthly rental Actors entirely, explaining plainly that "an AI agent cannot effectively manage a dozen different $50/month subscriptions." One reviewer's split verdict: 4.5/5 for developer pipelines, 2.0/5 for B2B sales prospecting.

Automation · Web Scraping · 4.4 ★

What is Apify?

Apify is a cloud-based web scraping and automation platform built around "Actors" — serverless cloud programs, each handling a specific scraping or automation task, run through a web UI, REST API, JavaScript/Python SDK, or on a schedule. Its Actor Store is the platform's standout feature: with 15,000-25,000+ pre-built Actors depending on the count cited, it's the largest actor marketplace in the web scraping space, covering everything from Amazon and Google Maps to social media, price tracking, and lead generation, each with an auto-generated UI so you can configure and run one without writing any code. For developers wanting more, custom Actors are fully buildable, and the platform integrates with Zapier and Make, both covered elsewhere in this directory, plus Google Suite, GitHub, and AI frameworks like LangChain and LlamaIndex. It's genuinely worth being precise about where Apify actually excels: one detailed 2026 review gives it a split rating of 4.5 out of 5 for developer data pipelines, but only 2.0 out of 5 for B2B sales prospecting specifically — a fair, honest scope clarification, since Apify functions as a data-sourcing layer rather than a full outbound sales stack, and teams evaluating it purely as a prospecting tool are likely to be disappointed relative to what a dedicated outbound platform delivers.

The genuinely major, current story worth understanding about Apify's 2026 business model evolution: in April 2026, the platform began formally sunsetting its "rental Actor" pricing model — fixed monthly rental fees charged by third-party Actor developers — in favor of a granular pay-per-event (PPE) system, with new rental Actors and pricing changes blocked from April 1 and full retirement of remaining rental Actors scheduled for October 1. Apify's own blog frames the shift directly around the rise of AI agents: "an AI agent cannot effectively manage a dozen different $50/month subscriptions; it needs to 'pay-as-it-goes.'" More than 2,000 Actors had already migrated from pay-per-result to pay-per-event pricing by that April announcement — a genuinely current, concrete example of AI agents reshaping how SaaS marketplaces themselves get billed. The core technical billing unit remains the Compute Unit (CU) — one CU equals one gigabyte of RAM running for one hour, with most scraping runs consuming just 0.01-0.5 CUs, though JavaScript-heavy Actors using headless Chrome cost 5-10 times more than lightweight, simple HTML-extraction Actors. The genuinely important thing worth understanding about the real bill you'll actually pay: it has two layers most first-time buyers miss. Layer one is your monthly plan fee plus its matching prepaid platform-usage credit; layer two is per-Actor pricing, since many Store Actors charge per result, per event, or a monthly rental fee on top of whatever compute your run consumed — meaning you can genuinely blow through a $29 monthly budget on a single pay-per-result Actor without the platform's own usage chart ever looking alarming.

🔄
Rental Actors retiring by October 2026
Replaced by pay-per-event pricing, explicitly framed around AI agent economics
⚠️
Two-layer billing most buyers miss
Plan credit + separate per-Actor fees that can blow past your budget invisibly
⚖️
4.5/5 for data pipelines, 2.0/5 for B2B sales
A fair, honest scope split — data-sourcing layer, not a full outbound stack
💻
JS-heavy Actors cost 5-10x more
Headless Chrome runs cost meaningfully more than simple HTML extraction

The April 2026 rental-model retirement and its AI-agent framing are drawn directly from Apify's own official blog announcement. The two-layer billing gotcha and CU-based cost math are drawn from a detailed independent 2026 pricing guide (Use Apify), verified against Apify's official pricing page. The split rating for developer pipelines versus B2B sales prospecting is drawn from an independent comparative review (SyncGTM).

Key features

🛒

15,000-25,000+ Actor marketplace

The largest pre-built scraper library in the category, with auto-generated UIs.

💻

Full developer SDK & API

Build custom Actors in JavaScript or Python, or run existing ones programmatically.

🤖

MCP server integration

Genuinely praised support for AI-native workflows via Model Context Protocol.

⚙️

Pay-per-event pricing (2026)

Actors bill for specific actions, not broad, unpredictable outcomes.

🔌

Zapier, Make, GitHub integrations

Schedule and connect Actors directly into existing automation workflows.

🧠

LangChain & LlamaIndex support

Feed scraped data directly into LLM-powered applications and agents.

Available models

Apify MCP Server Exposes the Actor Store to AI agents for programmatic, pay-per-event data fetching

Integrations & platforms

Zapier, Make, Google Suite LangChain, LlamaIndex REST API, JS/Python SDK

Pros, cons & best for

👍

Pros

  • Genuinely the largest, most developer-friendly Actor marketplace in the category
  • Real, current AI-agent-native positioning with strong MCP server support
  • Full SDK/API flexibility for custom, complex scraping and automation logic
👎

Cons

  • Two-layer billing (plan credit plus per-Actor fees) genuinely catches first-time buyers off guard
  • Weak fit specifically for B2B sales prospecting versus dedicated outbound tools
  • JavaScript-heavy Actors cost meaningfully more, a real budgeting consideration
🎯

Best for

  • Developers and data engineers building genuine scraping/data pipelines
  • Teams building AI agents that need programmatic, pay-per-use web data access
  • Not the pick as a full B2B sales prospecting or outbound stack

Take a look inside

Our verdict

4.4 / 5

Apify's genuine, category-defining strength is real developer flexibility at real scale — the largest Actor marketplace available, full SDK and API access, and genuinely strong MCP server support position it well for the current shift toward AI agents that need to fetch web data programmatically, a shift the platform's own April 2026 retirement of fixed rental pricing directly reflects. The honest, important thing worth understanding before committing budget: the real bill has two layers, and many first-time buyers genuinely underestimate how fast per-Actor fees stack on top of the base platform credit, especially with JavaScript-heavy Actors running 5-10x pricier than simple ones. It's also fair to be precise about scope — this is a genuinely strong data-sourcing platform for developers, not a full B2B sales prospecting stack, and evaluating it as the latter will likely disappoint. For developers and data engineers building real scraping pipelines or AI-agent data infrastructure, Apify remains one of the strongest, most flexible choices available; for B2B sales prospecting specifically, a dedicated outbound tool is the better fit.

FAQ

Why is Apify retiring its rental Actor pricing model?

Starting April 2026, Apify began sunsetting fixed monthly rental fees in favor of pay-per-event pricing, explicitly framing the shift around AI agents: an autonomous agent can't effectively manage a dozen different $50/month subscriptions, so the platform needed to become genuinely pay-as-you-go instead.

Why did my Apify bill end up higher than expected?

Because the real bill has two layers most first-time buyers miss — your monthly plan fee with its matching platform-usage credit, plus separate per-Actor pricing that many Store Actors charge on top of the compute your run actually consumed.

Is Apify good for B2B sales prospecting?

Not especially — one detailed review rates it just 2.0 out of 5 for that specific use case, noting Apify is a data-sourcing layer rather than a full outbound sales stack, compared to a strong 4.5 out of 5 rating for developer data pipelines.

How is Apify's usage actually billed?

By Compute Units (CUs), where one CU equals one gigabyte of RAM running for one hour — most scraping runs consume just 0.01-0.5 CUs, though JavaScript-heavy Actors using headless Chrome cost 5-10 times more than simple HTML-extraction Actors.

Do unused Apify credits roll over to the next month?

No, monthly credits expire at the end of each billing cycle and do not carry forward, regardless of how much of your allowance went unused.

How can I estimate my real Apify costs before committing?

Run a small sample of 10-50 records first, check the actual Compute Units consumed in the Run detail view, extrapolate to your expected full volume, and add a 2x safety buffer for a realistic estimate.