Change log

New access rights system allows you to provide other users with limited access to your actors, tasks, and storages. Read more about access rights in blog post.

Apify’s documentation has been completely redesigned and now has its own domain at docs.apify.com.

Offers has been launched — a feature on Apify Marketplace that allows developers to prepare proposals for projects and share them directly with customers. See blog post for more information.

Each of storage types (key-value store, dataset and request queue) can be now named and renamed using API or manually in Apify app.

Dataset get items API endpoint now returns unlimited number of items.

Legacy Apify Crawler product has been retired in favor of the new apify/legacy-phantomjs-crawler actor running on the Apify Actors platform. For more information read this blog post.

See how web scraping and automation with Apify can help your business at use cases.

You can now run your Scrapy spiders at Apify Actors platform using the new apify/scrapy-executor actor.

Get inspired by awesome projects built using Apify at case studies.

Increased stability and lifetime of residential proxy sessions.

Apify CLI now deploys actors using multi-file source type to make the code readable in Apify app.

New help center is available at help.apify.com. It provides search across all the resources such as knowledge base articles, blog, documentation, Stack Overflow and Apify SDK website.

Dataset fetched in RSS format now contains <lastBuildDate> based on its modification date.

Apify Marketplace is now open for developers to join and earn money by developing custom solutions for Apify customers.

Apify website apify.com has been completely redesigned.

Page analyzer improved. Check it out!.

Finished (failed) actor run can be now resurrected back to a RUNNING state. Checkout out the documentation for more information.

Scheduler now displays log of both successfull and failed invocations.

Starting on Wednesday 31st July 2019, user accounts with disposable (temporary) email addresses will no longer be able to run actors, use Apify Proxy or call Apify API. To retain your access to Apify services, please change your email address on the Account - Settings page to a legitimate email provider.

Actor and Task inputs specified through INPUT_SCHEMA.json can now be split into collapsible sections. Find out how to configure them here.

Apify Crawler is being phased out. Please read this blog post to find out why we are retiring the Crawler product, what it means for you and how you can migrate your crawlers to a new actor, including the integrations.

Actor task input can be now retrieved and modified using a special API endpoint.

Users can now upload custom profile picture in account settings.

Users can upload image to published actor. This image will soon appear at its public page in Apify Store.

Startup times of actors were optimized using CPU boost during the first 10 seconds of run.

Actor run along with its data can be now shared using public link that is available under the tab "info".

Tasks now support only JSON encoded input. This also affects API which returns actor task input directly as object under the input property instead of JSON-encoded pair of body and contentType. See API documentation of a get actor task endpoint.

Apify Forum was moved to Stack Overflow.

Actor task input can be overloaded in scheduler.

Added limit of 300 characters for description.

New Dockerfile templates for multifile allow faster builds.

Rate limit for dataset push items endpoint increased to 300 req/s per store.

Added actor author username to list actor tasks.

Added input schema to build detail.

Schedules that use a predefined CRON expression such as @monthly, @weekly, @daily or @hourly randomly change the base times to ensure that schedules with the same expression will not all start at the same time. This measure is aimed at improving startup times and the performance of your actors and crawlers.

Tasks can be now easily published as actor. Check out knowledge base article to learn more.

Request payload can be now modified in webhook configuration. Check out webhooks documentation to learn more.

Ad hoc webhooks now support idempotency key to ensure that duplicate webhooks won't get created when actor gets restarted. Check out webhooks documentation to learn more.

Web server running in actor is not required to start in 120sbut can start at any time during the lifespan of its container.

Git deployment key is now available via API (get actor endpoint).

"Use spare CPU capacity" configuration was removed.

Per hostname traffic statistics are now shown at Apify Proxy page in app.

Data transfer is now limited to 5GB per IP address.

Improved performance of a Google SERP Proxy.

Two new generic scrapers released – apify/puppeteer-scraper and apify/cheerio-scraper.

Increased maximum memory for actor runs to 32 GB.

Input UI for actor now validates proxy configuration.

Added set of API endpoints to manage webhooks and retrieve webhook dispatches.

New validation options added to actor input schema field definitions. For example max and min length of string and array field or regular expression pattern for values of string list field.

URL of Apify proxy now supports new parameter country that restricts proxy IPs selection to given country.

Run can now metamorph into run of another actor.

Original Apify crawler has been open sourced as actor apify/legacy-phantomjs-crawler. This actor has the same input as original Apify crawler and also the same output format.

New set of API endpoints to retrieve and manage the last actor (task) run and its default storages. Check API documentation for more information.

Source code editor extended with multifile support, more in documentation.

Runs with READY and RUNNING state are now pinned to the top of the actor runs list.

New input UI fields added (key-value pairs, string list, hidden fields). All the field types now support nullable option. See documentation page for more information.

Improved actor publication page.

New webhooks component enables integration of actors with external services and orchestration of multiple actors into single pipeline.

Run console was improved and provides quick overview of actor run storages.

Published actors have new title that is displayed at its public library page.

Added support for hidden fields (i.e. fields starting with the # character). These fields may be used to store debug information such as errors, response codes, etc. that might be easily omitted from output.

Dataset: Added new parameters to API endpoints returning dataset items - skipEmpty=true to omit empty items, skipHidden=1 to omit hidden fields and clean=true a shortcut for skipEmpty and skipHidden.

All endpoints with [username]~[resourceName] parameter in URL now support also [userId]~[resourceName] format.

Code editor used at Apify app was replaced with modern Monaco editor that supports all ES6 features.

Memory limit for free accounts increased to 8GB.

Input UI for request list now supports web hosted or uploaded file with a list of URLs. Try out Crawler - cheerio to see it in action (Start URLs field).

Publicated actor can be now marked as deprecated. Deprecated actor will be omitted from public library search and flagged as deprecated. Use this feature to tell people your actor is no longer being developed, since removing it might break integrations that depend on the actor.

Replaced code editor with Monaco Editor which supports modern JavaScript features and provides better coding experience.

Removed the meta.clientIp field from several API endpoints due to privacy concerns

Updated base Apify Docker images to use CMD rather than ENTRYPOINT instruction to launch the code. If you're using a custom Dockerfile that is based on Apify base images, make sure your CMD instruction is correct. See Dockerfile example for more information.

Added featured actors and crawlers to library. Added input schema and example run to actor detail page.

Added new section with third-party login services to Account page

Numerous performance and stability improvements, bugfixes

Dataset detail page now shows preview of the data.

Added new commands to manage secrets environment variables, check apify secrets help for more details.

Simplified apify.json file structure. It will be updated automatically before execution apify run and apify push command. Read more in the documentation.

Added new Orders section to enable customers to keep track of their custom projects. Read more in a blog post.

A large number of user interface and performance improvements.

Now you can set an additional billing email address that will receive copy of all invoices. To set it, just go to your Subscription page, click Edit, set Billing email and click Update subscription.

Apify Storage API endpoints (i.e. key-value stores, datasets, and request queues) that use other than GET HTTP method are now authorized using API token of user. Please see API documentation for more information. Note that we made a special exception in the system that will ensure that affected users will be able to continue using the API the old way. We'll send additional information to these users.

New endpoints providing access to particular version of actor added.

Actor task input can be now overloaded via API. See documentation for more information.

Private Git repositories are now supported. Check documentation for more information.

Improved actor UI - run console and source page has been redesigned for better developer experience.

Improved search in library.

A new page with awesome case studies was published.

Actors and crawlers in library are now organized by categories.

"Is exclusive" functionality of scheduler now supports actor also. If this options is checked then scheduler won't start another run as long as previous is still running.

New documentation for Apify SDK is now available at https://sdk.apify.com.

Input of an actor and its input UI can be now described in input schema.

Many new public actors with UI for input released in library: petr_cermak/booking-hotels, lukaskrivka/google-spreadsheet, jakubbalada/content-checker, apify/image-diff, ... checkout library for more.

Released Apify actor tasks. Using them, you can create multiple configurations of a single actor and then run the selected configuration directly from Apify Platform, schedule or API.

New documentation of Apify proxy released. Contains examples in multiple languages and detailed description of all provided proxies - datacenter, residential, and Google SERP.

Released new major version v0.7 of apify NPM package. Check changelog for more information.

Changed behaviour of apify run command and apify local storage directory name. Check migration guide if you are updating from version v0.1.*.

Added actor live view that enables connecting to running containers - read more on Apify Blog

Major internal code consolidation and performance improvements

Various bugfixes and improvements in code and documentation

Improvements in Google SERP proxies, adding additional providers

Added support for input file from other steps.

Memory option for actor runs now supports only values that are power of 2 (ie. 128MB, 256MB, 512MB, 1024MB, 2048MB, ...)!

Proxy configuration of crawler now offers "automatic" mode that rotates all the proxies available for a user.

Each actor run can now start a web server accessible at a certain unique URL. This enables you to run a web server inside the actor to provide real-time snapshots or receive tasks on the fly. See documentation for more details.

Added API endpoints to abort Actor run and build.

New Apify Proxy service launched!

Added support for running actors in Keboola integration. Check knowledge article for more information.

Minimum memory for actor runs is now 128MB.

Added log streaming for apify push and apify call commands.

Added parameter to clean stores before runs actor locally. Check doc for more information.

Bunch of improvements and new features. Check the changelog.

Now it is not possible to combine custom proxies and Apify proxy groups.

Run console now shows information about current/max/avegare CPU and memory.

Actors are now notified 120s before migration to another worker machine. Check documentation for more information.

Added a new API end-point to obtain information about a user account

Storage API now also supports use of [username]~[storage-name] instead of Dataset ID and Key-value store ID.

We have just released an Apify CLI (command line tool) to simplify local development, debugging and deployment to Apify.

New storage type for Actor platform that helps to manage dynamic queue of URLs to be processed. Check storage documentation for more information.

apify NPM package contains a lot of new features. Check its changelog for details.

limit for number of processes per actor run was increased to 2 x [memory megabytes] so with 2 GB memory your limit is 4000 processes.

host machine now sends migrating event to actor process in a case of upcoming restart or shutdown. Check documentation.

actor runs have now fixed amount of CPU capacity reserved and therefore each run should take about the same time. We also added a new checkbox "Use spare CPU capacity" in actor settings allowing actors to use spare CPU capacity at host machine as free boost.

We released a new version of our open souce apify npm package containing a lot of new stuff to help you with your web scraping and automation projects. Check its npm page, source code at GitHub repository and the documentation.

apify/actor-node-puppeteer Docker image is now deprecated. Use apify/actor-node-chrome image instead.

We have added apify/actor-node-chrome-xvfb image that supports non-headless Chrome. If you choose this image then Apify.launchPuppeteer() opens Puppeteer with non-headless Chrome by default.

Method datasets.getItems() now returns object PaginationList with items wrapped inside instead of plain items array. This helps to iterate through all the items using pagination. This change is not backward compatible!

We did improvements of our infrastructure to improve actor starts and overall performance.

Logs are now rate-limited. Each actor run and build has 10 000 lines log credit with 10 lines added each second. Log lines over the limit won't be available in both UI and API.

Launched Page Analyzer tool to enable setting up crawlers with less manual steps. Read more on Apify blog.

Major improvements to our Linux server configuration to improve stability and performance of the system

Actors can now run with 16GB memory (available for users with Medium and large plans see https://apify.com/docs/actor#limits

Actor runs and their default key-value stores and datasets are now being deleted after data retention period.

We've added support for PayPal payments for all subscription plans

The actor source code can now come from a GitHub Gist, which is much simpler than having a full Git repository (read the docs)

We have re-launched the Knowledge base with a new design and much better search options.

Added API endpoint to run an actor and get its output in a single HTTP request.

We've added a new storage type Dataset. This enables you to store results in a way similar to Apify Crawler.

Actor usage statistics are now available in user account.

Smarter allocation of tasks to servers to improve performance

Environment variables can now also be passed to actor builds (as docker --build-arg parameter)

Added option to automatically restart actor runs on error

Fixed URL in the link element of RSS formatted last crawler execution result. This bug was causing that some RSS readers never refreshed the data

Added support for automatic rotation of user agents

Released a new NPM package called proxy-chain to support usage of proxies with password from headless Chrome

Added support for XLSX output Format for crawler results

Upgraded the web app to Meteor 1.6 and thus greatly improved the speed of the app

Improved internal notifications, performance and infrastructure improvements

Added feature to enable actor to be anonymously runnable

Apifier is dead, long live Apify! On 9th October we launched our biggest upgrade yet.

The old website at www.apifier.com was replaced with public static website www.apify.com and the app running at my.apify.com

A new product called Actor was introduced. Read more in our blog

Added actor support to scheduler.

Git and Zip file source type added to actor.

Account section has been completely revamped and uses tabs

Schedules section has been completely revamped and uses tabs

API endpoint to get crawlers results now accepts the skipFailedPages parameter.

API endpoint to get crawlers results now accepts the hideUrl parameter.

API endpoint providing results in XML format now allows to set XML tag names.

Added support for JSONL output format

Created Crawler request form to help customers specify the crawlers they would like to have built

Added finish webhook data feature that enables sending of additional info in webhook request payload. (see docs)

Added a feature to delete user account

Improvements in logging system

Officially launched Zapier integration

Added a new context.actId property that enables users to fetch information about their crawler. (see docs)

Consolidated logging in the web application, improvements in Admin interface

Added proxy groups crawler setting to simplify usage of proxy servers (see docs).

Added Schedule button to the crawler details page to simplify scheduling of the crawlers

Improvements in administration interface

Performance optimizations in UI

Added a tool to test the crawler on a single URL only (see Run console on the crawler details page)

Improved reports in admin section

Changed Twitter handle from @ApifierInfo to @apifier.

Bugfix - cookies set in the last page function were not persisted

Deployed some upgrades in data storage infrastructure to improve performance and reduce costs

Added sorting to Community crawlers.

Bugfixes, performance and cosmetic improvements.

improvements in administration interface.

Extended public user profile pages in Community crawlers.

Bugfix in exports of results in XML format.

Added a new context.actExecutionId property that enables users to stop crawler during its execution, fetch results etc. (see docs).

Improvements in internal administration interface.

Launched an external Apifier status page page to keep our users informed about system status and potential outages.

Numerous improvements on Community crawlers page, added user profile page, enabled anonymous sharing

Improved sorting of columns in CSV/HTML results table - values are now sorted according to numerical indexes (e.g. "val/0", ..., "val/9", "val/10")

Launched Apifier community page

Invoices are now in the PDF format and are sent to customers by email

We didn't launch anything today, just wishing you a happy Valentine's Day

New testimonials from ePojisteni.cz and Finbox.io published on our Customers page. Thanks Dušan and Andy!

Released a major upgrade of billing and invoicing infrastructure to support European value-added tax (VAT)

Added a new Video tutorials page

Improved normalization of URLs which is used by the crawler to determine whether a page has already been visited (see Request.uniqueKey property in docs for more details)

Changed CDN provider from CloudFlare to AWS CloudFront to improve performance of web and API

Bugfix in the start execution API endpoint - synchronous wait would sometimes time out after 60 seconds

further improvements in administration interface

improved aggregation of usage statistics, now it refreshes automatically

Request.proxy is now available even inside of the page function

improved Invoices page

improvements in administration interface

displaying snapshot of the crawling queue in the Run console

all paginated API endpoints now support desc=1 query parameter to sort records in descending order

added support for XML attributes in results

added support for RSS output format to enable creating RSS feeds for any website

launched a new discussion forum

custom proxy used by a particular request is now saved in Request.proxy field (see Custom proxies in docs)

performance improvements

enabled rate limiting

Major API upgrades

added new endpoints to update and delete crawlers

support for synchronous execution of crawlers

all endpoints that return lists now support pagination

API Reference was greatly improved

Added new Tag and Do not start crawler if previous still running settings to schedules

Added new Initial cookies setting to enable users to edit cookies used by their crawlers

Added a list of invoices to Account page

Added a new usage stats chart to Account page

Large improvements in the deployment system completed

Increased the length limit for Start URLs to 2000 characters

Showing more relevant statistics in crawler progress bar

Released a new shiny API reference

Performance and usability improvements in admin interface

Migrated our main database to MongoDB 3.2, deployed new integration test suite, new metrics in admin interface

Showing current service limits on the Account page, various internal improvements in user handling code

Added new example crawlers to demonstrate how to use page's internal JavaScript variable and AJAX calls

Released Schedules that enable to automatically run crawlers at certain times.

Switched to Intercom to manage communication with our users

Added functionality to test finish webhooks

Security fix - added rel="noopener" to all external links in order to avoid exploitation of the window.opener

Displaying Internal ID field on crawler details page, and User ID and API token token on the Account page to simplify setup of integrations

Added a new Jobs page, because we're hiring!

Deployed various performance optimizations and bugfixes

Updated our Meteor application to use ES2015 modules

Published a new testimonial from Shopwings on our Customers page. Thanks Guillaume!

queuePosition can now also be overridden in interceptRequest function (see docs)

Performance improvements of results exports

Added new example crawler to demonstrate a basic SEO analysis tool

Upgraded Meteor platform from version 1.3 to 1.4

Added API property name and type next to each crawler settings (see docs)

Added a new context.stats property to pass statistics from the current crawler to user code (see docs).

Added a new signature for context.enqueuePage() function that enables placing new pages to beginning of the crawling queue and overriding uniqueKey and label fields (see docs).

Enabled users to define custom User-Agent HTTP header, updated the default value to resemble latest Chrome on Windows.

Implemented optimization that enables user to export even large result sets to CSV/HTML format.

Created this wonderful page to keep our users up-to-date with new features