The Apify web scraping, data extraction, and automation platform

What do you like best about the product?

What I like most is the speed to get a working scraper running. Prebuilt actors plus built-in infrastructure (proxies, scheduling, scaling) remove most of the setup work we used to handle manually.

What do you dislike about the product?

The main downsides are the learning curve for custom Actors and the cost becoming less predictable at scale (especially with heavy runs, proxies, and multiple workflows).

What problems is the product solving and how is that benefiting you?

Apify solves the problem of building and maintaining web scrapers in-house.
Instead of spending time writing and constantly fixing scraping infrastructure, we can rely on managed actors and built-in scaling. This reduces engineering effort, speeds up data collection, and makes scraping pipelines more stable.

What do you like best about the product?

Apify has an incredibly robust network of apps and scrapers to choose from. Its ease of use, affordability, and the simplicity of setting up integrations feel unparalleled, which helps drive successful deployments and consistently strong performance.

What do you dislike about the product?

The inconsistency of the pricing between the individual products. It feels quite arbitrary

What problems is the product solving and how is that benefiting you?

We use Apify constantly. The depth and breadth of the service offerings helps us create agents and bots with relative ease. The ability to test and iterate to ensure strong performance is a blessing and overall learning curve is quite slim.

What do you like best about the product?

Good UX and diverse integrations pretty good price and performance

What do you dislike about the product?

sometimes the output is not super relevant and the actors are not super customizable

What problems is the product solving and how is that benefiting you?

scraping potential leads for my business

What do you like best about the product?

The Actor marketplace is incredibly valuable. I'm building an automated lead generation system using AI (Claude API for scoring, Playwright for automation) and Apify handles all the web scraping infrastructure. Pre-built Instagram Actors give me structured data that I pipe directly into my AI scoring pipeline. The REST API integration with my Node.js/Express backend was seamless — well-documented endpoints, predictable response formats, and solid error handling.

What do you dislike about the product?

The learning curve for building custom Actors is steeper than I expected. Also, some third-party Actors in the marketplace vary in quality — it would help to have more standardized output schemas across similar Actors.

What problems is the product solving and how is that benefiting you?

Before Apify, I was spending 60-70% of my development time on scraping infrastructure — proxy rotation, browser fingerprinting, anti-bot evasion, and data parsing. Now that's all handled by the platform, and I can focus on the AI and business logic of my application. It's reduced my time-to-market significantly and the per-usage pricing keeps costs predictable for a bootstrapped project.

What do you like best about the product?

The MCP gave me the ability to build and deploy 7 actors collecting 250,000+ rows of marketplace data in 6 weeks — by talking to Claude. The core primitives (call-actor, get-actor-run, get-actor-output) are the right shape for agentic workflows. search-actors and fetch-actor-details let me discover existing actors instead of reinventing the wheel. And search-apify-docs is underrated — having Claude pull up the right docs without context-switching to a browser is a small thing that compounds.
The bigger insight: Apify composes well with other MCPs. In my workflow it's Apify (collect) + Supabase (store) + Claude (orchestrate). Each does one thing; the conversation stitches them together. That multi-MCP pattern is where I think the real value is.

What do you dislike about the product?

The default Claude desktop / Cowork connector ships with a fairly limited tool set. The core operational tools (run history, abort, logs, schedule listing) do exist, but they’re tucked away in a separate MCP Configurator—and I didn’t even know the configurator existed until a PM personally walked me through it. Up to that point, I assumed those capabilities were missing entirely and ended up falling back to REST API workarounds.

This feels more like a discoverability problem than an actual product gap, but it still cost real time. Either ship the operations tools enabled by default, or make the configurator clearly visible during connector setup so it’s obvious where those features live.

What problems is the product solving and how is that benefiting you?

I’m building a market intelligence database for vintage synthesizers by pulling pricing history from eBay, Reverb, VEMIA, Yahoo Japan, and other marketplaces. It’s something I’ve wanted to do for years, but I couldn’t make it happen—the technical lift was simply too high for someone who hasn’t written production code in over a decade.

What changed for me is a new generation of enablers, and Apify is one of the main ones. The platform gave me both the runtime to deploy scrapers and the MCP to orchestrate them through natural language. As a result, what would have been a multi-month engineering project turned into a few weeks of conversation-driven setup: 7 deployed actors and 250,000+ entries collected in about 6 weeks.

Overall, the combination of Apify’s reach and the MCP’s accessibility has unlocked ambitions that used to feel locked behind “you need to be a developer.”