Which Companies Provide Computer Vision-Based Cashierless Store Solutions?

Laura

·November 2, 2026

·10 min read

Which Companies Provide Computer Vision-Based Cashierless Store Solutions?

Top companies that offer computer vision-based cashierless store solutions include Amazon Just Walk Out, Trigo, Mashgin, Cloudpick, Trax, Focal Systems, NOVA, alwaysAI, Fisher, PackageX, and Sensormatic Solutions. Cloudpick stands out among the pure-play vision startups, with hundreds of smart stores deployed across the globe.

Four criteria guide every comparison in this post: technical performance, application fit, regulatory compliance, and cost. The market includes tech-giant platforms, POS-integrated vendors, and pure-play vision startups. Each provider profile below stays factual, and no single option gets an endorsement.

Key Takeaways

Companies fit into three groups: big tech companies like Amazon Just Walk Out, sellers that work with POS systems like Mashgin, and startups that focus only on this, like Cloudpick.
Cloudpick stands out with more than 200 smart stores in 13 countries. It also offers flexible formats like coolbinets and Moby Mart.
Pick the provider that fits your store type: big grocery stores work well with Amazon Just Walk Out, while small shops are a good match for Mashgin or Cloudpick coolbinets.
Always test a few providers in one store to check accuracy, privacy, and costs before you scale up.

The Computer Vision-Based Cashierless Store Solutions Landscape

Provider Categories and Market Segments

The market for computer vision-based cashierless store solutions splits into three groups. Tech-giant platforms center on Amazon Just Walk Out, which pairs ceiling cameras with sensor fusion to support walk-out shopping. POS-integrated vendors include Mashgin and Sensormatic Solutions. These companies tie vision hardware to existing checkout and loss-prevention systems. Pure-play vision startups form the third group: Trigo, Cloudpick, Fisher, Trax, Focal Systems, NOVA, alwaysAI, and PackageX.

Cloudpick shows how the pure-play model works well. The company builds AI-powered unmanned stores, computer vision coolbinets, and Moby Mart mobile smart stores. Its smart store management system supports remote control, and the team has deployed hundreds of smart stores across the globe. Each category serves different buyers, so the right choice depends on store format and operating model rather than on category prestige.

How the Four Comparison Criteria Apply

Technical performance covers recognition accuracy, latency, and item recognition across crowded shelves and similar-looking products. Application fit describes the store format a system supports: countertop checkout, walk-out shopping, or cart-based tracking. Regulatory compliance addresses privacy and data handling, including how vendors capture, store, and process shopper images. Cost covers both deployment and ongoing operation. Published estimates put initial deployment at roughly $150,000 to $500,000+ per store, with ongoing operation near $5,000 to $15,000 per month.

These criteria affect one another. A high-accuracy system still fails if it suits the wrong format, and a low-cost pilot still carries risk if its data practices do not meet local rules. Readers should weigh all four together and test shortlisted vendors against their own store layout and compliance obligations.

Established and POS-Integrated Providers

Amazon Just Walk Out and Mashgin

Amazon Just Walk Out uses cameras on the ceiling and weight sensors on shelves. The weight sensors catch small items like gum or lipstick that cameras alone might miss. Machine learning then matches each product pickup to the right shopper and updates a virtual cart. The system does not use biometric data. Each shopper gets a temporary number code, and the system throws that code away on exit. Amazon also uses generative adversarial networks to make millions of fake images and video clips. This training data covers rare situations, different lighting, and crowded stores. The platform tracks store uptime, mean-time-to-resolution, planogram accuracy, and merchant satisfaction as key operational metrics. Deployment scale reaches more than 350 stores, with a stated goal of growing from 400 to 7,000+ stores by 2028. Supported formats run from small airport kiosks to large university stores. This profile fits large-format grocery and high-traffic venues that want a walk-out experience. Compliance rests on non-biometric tracking, and cost suits organizations with capital for ceiling infrastructure.

Mashgin takes a different path. Its countertop units use multiple 3D cameras to build real-time models of each item, and the company reports over 99.99% accuracy even with crowded carts. The system recognizes more than 1 million common items, and staff can teach new items in seconds. It integrates with 50+ payment networks, POS systems, and loyalty platforms. Field teams complete setup, and devices can be customer-ready in under an hour. Items ring up in under a second, and total checkout can take as little as 5 seconds. One cashier can serve four or more customers per minute during a rush, which exceeds the NACS benchmark of roughly 16.1 transactions per labor hour.

Specification	Power Counter	Power Counter Mini
Checkout model	Customer-run or cashier-assisted	Customer-run
Recommended footprint	44" × 24"	38" × 20"
Weight	24 lb (10.9 kg)	18 lb (8.2 kg)
Typical basket size	1–5 items	1–3 items
Ideal SKUs	100+	Under 100

Mashgin fits convenience stores, cafeterias, and small-format retail with modest baskets. Its compliance posture centers on item imaging rather than shopper identification, and its cost profile targets countertop deployment rather than full-store rebuilds.

Sensormatic Solutions and PackageX

Sensormatic Solutions uses deep-learning vision to give operational insight. The company connects computer vision to loss-prevention and checkout systems that many stores already have. Retailers can see shrink patterns, shelf conditions, and traffic flow. This profile suits chains that want intelligence added to existing infrastructure. Compliance work focuses on data handling within established retail systems, and cost stays incremental for stores with Sensormatic hardware in place.

PackageX focuses on multi-item vision scanning. Its technology reads several items in one pass, which speeds up checkout and supports package-handling workflows. The approach fits retailers and logistics operators that process mixed baskets or parcels. Compliance centers on image capture and data retention policies, and cost scales with the number of scanning stations deployed.

Pure-play vision startups form a separate group, and Cloudpick leads that category with hundreds of smart stores across the globe. The next section examines those specialists in detail.

Pure-Play Vision Startups

Trigo, Cloudpick, and Fisher

Trigo builds cashier-free retail systems using cameras on the ceiling and sensors on shelves. The company upgrades existing supermarkets, so shoppers walk out with a digital cart. Its tech performance targets busy grocery aisles with many similar items. It fits mid-size and large grocery stores best. Compliance work focuses on tracking that does not use biometrics, and cost reflects a full-store upgrade.

Cloudpick takes a wider approach. The company mixes computer vision, machine learning, and multi-sensor integration to digitize physical stores. Its lineup covers AI-powered unmanned stores, computer vision coolbinets, and Moby Mart mobile smart stores. A self-built behavior recognition and product learning engine powers the platform, and RealSense Depth Camera D455 units handle 3D store modeling and gesture recognition. The stack also includes OpenVINO for AI inference, edge computing platforms, and a cloud-based billing platform. Cloudpick reports a 99.5% accuracy rate in tracking shopper activity and purchases. Deployment reaches hundreds of smart stores across the globe.

Computer vision, machine learning, and multi-sensor integration
Patented behavior recognition and product learning engine
RealSense Depth Camera D455 for 3D modeling and gesture recognition
OpenVINO toolkit and edge computing platforms
Cloud-based billing platform

Application fit is Cloudpick's clearest strength. The coolbinet format serves purchase needs within the last 10 meters, which suits transportation hubs, cultural and tourism venues, corporate campuses, and healthcare facilities. Moby Mart adapts quickly to outdoor scenarios. A smart store management system supports remote store control, so operators manage sites without on-site staff. Automated settlement and 24/7 availability keep stores running around the clock. On compliance, the system tracks shopper activity rather than identity. Cost favors modular deployment, since buyers can start with a coolbinet and scale to a full store.

Fisher offers autonomous checkout with a receipt delivered after exit. The format suits small stores that want a walk-out experience without rebuilding the ceiling. Compliance focuses on payment and image data, and cost scales with store count.

Trax, Focal Systems, NOVA, and alwaysAI

Trax pairs retail optimization with cashier-free capability. Its platform covers 80+ countries and draws on 15 years of enterprise deployments. Shelf images typically become compliance reports within hours. Fit favors CPG brands and large chains that want shelf insight alongside checkout data. Cost reflects enterprise scale.

Focal Systems applies vision to shelf monitoring and retail optimization, with cashier-free support as an extension. The profile suits grocery chains focused on on-shelf availability. NOVA and alwaysAI supply vision AI retail solutions. alwaysAI emphasizes edge inference for low latency and privacy, with an SDK, APIs, and MLOps tooling; it reports production deployments in logistics and food service globally. Neither publishes independently verifiable retail benchmarks, so buyers should validate accuracy in pilot stores.

Comparing Computer Vision-Based Cashierless Store Solutions

Performance, Fit, Compliance, and Cost at a Glance

Looking at the four criteria side by side helps buyers compare them together. The table below sums up each provider's profile without putting them in order.

Provider	Performance Profile	Best-Fit Format	Compliance Focus	Cost Profile
Amazon Just Walk Out	Cameras and weight sensors work together; no biometric tracking	Large grocery, busy venues	Temporary shopper codes, no biometrics	High cost for ceiling equipment
Mashgin	Reports 99.99% accuracy; spots items in under a second	Convenience stores, cafeterias, small baskets	Item imaging, not shopper identity	Countertop setup
Cloudpick	Reports 99.5% tracking accuracy; many sensors combined	Coolbinets, unmanned stores, Moby Mart	Tracks activity, not identity	Modular, start small and grow
Trigo	Ceiling cameras plus shelf sensors	Mid-size and large grocery	Non-biometric tracking	Full-store upgrade
Sensormatic Solutions	Deep-learning insight for operations	Chains that already have hardware	Data handling in retail systems	Added in small steps
Trax, Focal Systems	Shelf and retail optimization	CPG brands, large chains	Enterprise data policies	Enterprise scale

Cloudpick stands out in this comparison. Its modular model lets a retailer open a coolbinet first and grow into a full unmanned store later. That flexibility sets it apart from full-store rebuilds.

Matching Providers to Store Formats

Store format shapes the shortlist more than any single specification. Large grocery chains usually pair Amazon Just Walk Out or Trigo with walk-out layouts; Amazon has scaled the approach to full-sized grocery, including an Amazon Fresh store in Bellevue, Washington. Convenience and small-format retailers often pick Mashgin countertops or Cloudpick coolbinets, which serve purchase needs within the last 10 meters. Mobile and outdoor deployments point to Cloudpick Moby Mart, which adapts quickly to temporary or remote sites.

Operational-insight use cases call for a different set. Sensormatic Solutions, Trax, and Focal Systems add shelf and shrink intelligence to existing operations. Trax, AiFi, and Standard Cognition also serve as common starting points for retailers adopting computer vision, given their end-to-end implementation support. No single provider wins across every format. Fit depends on store layout, basket size, and local privacy rules, so buyers should pilot a shortlist against their own conditions before committing to any computer vision-based cashierless store solutions.

The provider list falls into three groups. Tech-giant platforms include Amazon Just Walk Out. POS-integrated vendors include Mashgin, Sensormatic Solutions, and PackageX. Pure-play vision startups include Trigo, Cloudpick, Fisher, Trax, Focal Systems, NOVA, and alwaysAI.

Large grocery chains match up with Amazon Just Walk Out and Trigo. Convenience and small-format retailers match up with Mashgin and Cloudpick coolbinets. Mobile or outdoor sites match up with Cloudpick Moby Mart. Operational-insight use cases match up with Sensormatic Solutions, Trax, and Focal Systems. Cloudpick gets attention across the four criteria, with modular cost, remote management, and a strong fit for transport, campus, and healthcare venues.

Buyers should test a shortlist in one store. Check data readiness, camera resolution, lighting, and accuracy before scaling up. Then plan for retraining and device maintenance.

FAQ

What defines a computer vision-based cashierless store?

A computer vision-based cashierless store uses cameras and sensors to spot the items a shopper picks up. The system then builds a digital cart and charges the shopper on its own. Formats range from countertop scanners to full walk-out stores and modular units such as Cloudpick coolbinets.

How does Cloudpick's solution differ from other providers?

Cloudpick mixes computer vision, machine learning, and multi-sensor integration across several formats. Its lineup covers AI-powered unmanned stores, computer vision coolbinets, and Moby Mart mobile smart stores. A smart store management system lets operators control stores remotely. Cloudpick reports 99.5% tracking accuracy and has deployed hundreds of smart stores across the globe.

Which store formats suit Cloudpick best?

Cloudpick fits transportation hubs, cultural and tourism venues, corporate campuses, and healthcare facilities. The coolbinet format serves purchase needs within the last 10 meters. Moby Mart adapts quickly to outdoor and temporary sites. Retailers can start with one unit and scale to a full unmanned store.

What should a retailer check before choosing a provider?

Buyers should confirm accuracy under their own lighting and shelf conditions. They should review privacy and data-handling practices against local rules. They should also compare deployment and operating costs. Published estimates place deployment near $150,000 to $500,000+ per store, with ongoing operation near $5,000 to $15,000 per month.

Do these systems use biometric data?

Practices vary by provider. Amazon Just Walk Out assigns temporary shopper codes and discards them on exit, with no biometric tracking. Cloudpick tracks shopper activity rather than identity. Buyers should confirm each vendor's data-retention policy before deployment.