Mobile Human-In-The-Loop

Collect And Label Machine Learning Data With Mobile Devices

Inspired By Amazon's Mechanical Turk
Built For The Mobile Era

Native mobile user interfaces Out of the box

The Swarms app turns smartphones and tablets into powerful tools for data collection and labelling. Our API provides a wide range of components which you can use to create your individual task layouts which work reliably on every iOS and Android device. Building own interfaces for your machine learning projects is a thing of the past. 

Very Fast Human Loops

Smartphones provides the fastest access to people with relevant skills for your tasks. Our platform can send push-notifications to activate qualified workers within seconds. With Swarms you can build very fast human loops which are especially helpful for validating the output of applied neural networks.  

Build Fully Automated pipelines with our API

Our API provides you all you need to build automated data collection and labelling pipelines. Use our API to qualify and manage workers, create multi-step processes (e.g. object classification, object localization, object segmentation, result validation by other workers), and review results. 

For more information check our API


Setup via API


We can assist you with setup if required.

Worker Payment

Starting from $0.01 per task

You offer a payment - workers can accept.

Swarms Commission

20-30% of worker payment

Minimum $0.01.

We provide commission discounts to NGOs, Researchers, Early Stage Startups.
Projects with a social mission can also publish unpaid tasks which people complete as a donation of their time.

Work with refugees - change lifes

We collaborate with the World Food Programme of the United Nations to provide refugees access to life-changing income opportunities. Our mobile platform links machine learning teams with refugees who collect and label machine learning data on their smartphone in exchange for payment. Refugees provide high quality results and achieve earnings far above the regional market average using Swarms.