What Amazon Mechanical Turk does

Amazon Mechanical Turk (MTurk) is a crowdsourcing marketplace operated by Amazon Mechanical Turk, Inc., a wholly-owned subsidiary of Amazon.com, Inc. Founded in 2005 and headquartered in Seattle, MTurk operates a two-sided platform that connects Requesters (businesses, researchers, and developers) with a distributed global workforce of Workers who complete Human Intelligence Tasks (HITs) — including image annotation, content moderation, data validation, transcription, survey participation, and research data collection — for micro-payments. The platform is accessible via a self-serve Requester web UI, a Worker portal, a full REST API exposed through AWS SDKs (Python, Java, JavaScript, .NET, PHP, Ruby, iOS, Android, Node.js), and a Developer Sandbox for testing, with a documented maximum of 250,000 HITs per batch and continuous 24x7 global availability across 35+ countries.

The core technology stack is built around a RESTful API with multiple SDKs, a qualification system (Masters and 100+ Premium Qualifications) for quality control, and direct integration with AWS SageMaker Ground Truth and Ground Truth Plus, which positions the MTurk workforce as an embedded data-labeling resource for ML pipelines including Generative AI use cases. Platform add-ons include customizable task templates across vision, language, and structured-data categories.

MTurk generates revenue through transaction fees: a 20% platform fee on Worker rewards and bonuses (with an additional 20% on HITs with 10+ assignments), a 5% premium for Masters Qualifications, and $0.05–$1.00+ per assignment for Premium Qualifications. Payment is processed through AWS billing, Prepaid Task Credits, or invoicing for enterprise Requesters. The go-to-market motion is entirely self-serve and developer-first, with no traditional sales team; distribution is amplified through AWS ecosystem integration and the SageMaker Ground Truth channel. Customer segments are horizontal and span machine learning/AI companies, academic and market research, content moderation, and data processing operations, with named enterprise users including Pinterest, Allen Institute for AI, Baidu Research, Johnson & Johnson, Maxar, US Foods, and Yum! Brands.

Name

Amazon Mechanical Turk

Legal name

Amazon Mechanical Turk, Inc.

Website

https://mturk.com

Company type

Public

Founded year

2005

Operating status

Operating

Headcount range

11–50 employees

Short description

Amazon Mechanical Turk is a crowdsourcing marketplace launched by Amazon in 2005 that connects Requesters with a global distributed workforce for completing Human Intelligence Tasks such as data labeling, annotation, moderation, and survey participation, monetized via a 20% transaction fee and embedded in AWS.

Ownership category

akta.pro rank

Product category

Crowdsourcing Marketplace

NAICS

Professional Employer Organizations (56133)

SIC

Services-Employment Agencies (7361), Services-Help Supply Services (7363)

akta.pro primary industry

Contractor / Freelancer Payments (FSAMAKAC)

Keywords

• Crowdsourcing marketplace
• Human intelligence tasks
• Data annotation services
• Microtask workforce
• Machine learning labeling

Headquarters

HQ city

Seattle

HQ country

United States

HQ region

North America

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Infrastructure, Operations, Personnel, Marketing or Sales

Revenue model

1. Transaction Fees (MTurk Fee): MTurk charges a 20% fee on the reward and bonus amount paid to Workers. For HITs with 10 or more assignments, an additional 20% fee applies on the reward. Minimum fee is $0.01 per assignment or bonus payment. This is the primary revenue stream based on transaction volume.
2. Masters Qualification Fee: Additional 5% fee on the reward paid to Workers when Requesters use the Masters Qualification to access top-performing verified Workers.
3. Premium Qualifications Fee: Additional per-assignment fees ranging from $0.05 to $1.00+ depending on the specific Premium Qualification used (demographics, financial, employment, location-based).

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels5 records

Core offering

Amazon Mechanical Turk (MTurk) is a two-sided crowdsourcing marketplace that connects business Requesters with a distributed global workforce of Workers who complete Human Intelligence Tasks (HITs) for per-task compensation. Requesters create and publish tasks such as data validation, image annotation, content moderation, survey participation, and audio transcription through a web-based UI, the MTurk REST API, or AWS SageMaker Ground Truth. MTurk charges a 20% platform fee on Worker rewards, with additional fees for the Masters Qualification (5%) and Premium Qualifications ($0.05 to $1.00+ per assignment).

Product overview

Amazon Mechanical Turk (MTurk) is a crowdsourcing marketplace that operates as a single, unified platform connecting businesses and researchers (Requesters) with a distributed global workforce (Workers) who complete Human Intelligence Tasks (HITs) for micro-payments. The platform consists of a web-based Requester UI with customizable task templates, a Worker portal for finding and completing tasks, a full REST API accessible via AWS SDKs (Python, Java, JavaScript, PHP, .NET, Ruby, iOS, Android, Node.js), and a Developer Sandbox for testing. Optional paid add-ons include the Masters Qualification for vetted top-performers (5% additional fee) and over 100 Premium Qualifications for demographic targeting ($0.05–$1.00 per assignment). MTurk also integrates with Amazon SageMaker Ground Truth, enabling managed ML data labeling workflows that leverage the MTurk workforce.

Differentiator

Problem solved

Functional benefit

Products and services

• Amazon Mechanical Turk (MTurk) A two-sided crowdsourcing marketplace that connects business Requesters with a distributed global workforce of Workers who complete Human Intelligence Tasks (HITs) — including data validation, image annotation, content moderation, survey participation, audio transcription, and research data collection — for per-task compensation. Provides web-based Requester and Worker UIs, a full REST API, qualification systems, and quality controls; transactions are settled with Workers receiving rewards and MTurk retaining a platform fee.
• MTurk Requester UI A web-based user interface that allows business Requesters to create, publish, and manage Human Intelligence Tasks without writing code. Includes customizable templates for surveys, image annotation, text processing, and other task types, batch CSV upload for publishing multiple HITs simultaneously, and direct access to qualification, approval/rejection, and result-retrieval workflows.
• MTurk API A RESTful web service API exposing MTurk functionality (createHIT, approveAssignment, rejectAssignment, etc.) via production endpoint https://mturk-requester.us-east-1.amazonaws.com and sandbox endpoint https://mturk-requester-sandbox.us-east-1.amazonaws.com. Supports integration through AWS SDKs for Python (boto3), Java, JavaScript, PHP, .NET, Ruby, iOS, Android, and Node.js, and integrates with AWS Identity and Access Management (IAM) credentials.
• MTurk Developer Sandbox A simulated MTurk environment mirroring the production service that allows developers to test applications and HITs before publishing to the live marketplace. Provides a static $10,000 test balance; no fees are charged and no real payments are made to Workers. Available separately as Requester (requestersandbox.mturk.com) and Worker (workersandbox.mturk.com) sandbox environments, with a 500-HIT per batch limit versus 250,000 in production.

Quantifiable outcome

• Transforms fixed hiring costs into variable pay-per-task costs
• +2 more outcomes

Named customers11 records

Segments4 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration3 records

AI capability8 records

Feature5 records

Partnerships

Two partnerships are on record, tiered core and minor.

• Amazon SageMaker Ground TruthcoreTechnology or IntegrationMTurk workforce is accessible through AWS SageMaker Ground Truth and Ground Truth Plus, providing integrated data labeling workflows for ML development. Ground Truth allows Requesters to implement additional validation and quality checks, with curated workforces for Generative AI use cases.
• Allen Institute for AIminorStrategic or Co-development PartnerAI2 uses Amazon Mechanical Turk to build datasets that help their models learn common sense knowledge for advancing AI research. MTurk provides flexible platform enabling harness of human knowledge for ML model training.

Scale indicators3 records

Recent moves6 records

Expansion highlights4 records

Market position

Strengths4 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

• Xx.com/amazonmturk
• GitHubgithub.com/awslabs/mturk-code-samples

Revenue estimate

Valuation estimate

Number of profiles

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Amazon Mechanical Turk