Top 10 US Computer Vision Companies

Scale AI reached a $7.3B valuation as the leading data annotation and AI evaluation platform, labeling 100M+ images and video frames annually for customers including the US Army, OpenAI, Microsoft, and GM. Its Nucleus platform enables teams to find failure modes in CV models and curate the most impactful training data. In 2025, Scale AI was awarded a $249M US Army contract for computer vision data labeling for autonomous vehicle and drone programs.

Landing AI, founded by Andrew Ng, reached an $850M valuation with its LandingLens visual AI platform enabling non-ML engineers to deploy industrial CV inspection without deep learning expertise. Its defect detection systems are deployed in semiconductor fabs, pharmaceutical manufacturing, and food processing, reducing inspection cycle times from hours to milliseconds. In 2025, Landing AI launched Visual Prompting technology that allows CV models to be fine-tuned from just 5-10 labeled examples — 99% less data than traditional approaches.

Cognex generated $800M in 2024 revenue holding 20%+ global machine vision market share with its In-Sight smart cameras and VisionPro software deployed in automotive, electronics, and pharmaceutical manufacturing. Its deep learning ViDi suite detects surface defects invisible to traditional rule-based vision systems. Cognex systems perform over 1 billion quality inspections daily worldwide, and its 2025 Edge Learning launch allows on-device model training on production lines without cloud connectivity.

Matterport generated $39M in Q1 2025 revenue from its 3D spatial intelligence platform with 11M+ scanned spaces spanning real estate, construction, insurance, and facilities management. Its Pro3 camera creates millimeter-accurate digital twins of physical spaces in under 60 minutes. In 2025, CoStar Group acquired Matterport for $1.6B, integrating its spatial scanning technology into the largest commercial real estate data platform and enabling automatic property condition assessment from digital twin data.

Roboflow raised $40M to build the developer platform for computer vision, serving 250,000+ developers and 10,000+ organizations who have trained 100,000+ CV models on the platform. Its dataset management, annotation tools, and model training pipeline compress CV application development from months to days. In 2025, Roboflow launched RF-DETR — its open-source real-time object detection model that outperforms YOLO on COCO benchmarks — becoming the most-starred CV model repository on GitHub within 90 days.

Labelbox reached a $1B valuation with its enterprise data labeling platform processing 100M+ labels per day for Fortune 500 clients including Procter & Gamble, Walmart, and DoorDash. Its Model-Assisted Labeling feature uses active learning to automatically pre-label images, cutting human annotation time by 70%+. In 2025, Labelbox launched Catalog — a semantic search engine for unstructured visual data that lets teams find specific CV training examples using natural language queries across billion-image datasets.

Clarifai raised $100M+ as an enterprise computer vision API platform serving food safety inspection, retail shelf analytics, and defense surveillance use cases. Its multi-modal AI platform processes images, video, text, and audio in a unified pipeline, enabling complex CV workflows like automatic shelf inventory analysis combining product recognition and OCR. In 2025, Clarifai deployed its CV platform with the US Department of Defense for automated aerial surveillance analysis, processing 10,000+ hours of drone footage monthly.

Wayve raised $1.05B in 2024 — the largest single investment in a UK AI company — from SoftBank, Microsoft, and NVIDIA to develop its Embodied AI approach to autonomous driving using imitation learning rather than rule-based systems. Its AV2.0 model learns to drive by watching humans, then generalizes to new roads without explicit programming. In 2025, Wayve launched US commercial pilots in San Francisco and Austin, and its foundation model technology was licensed to Amazon for autonomous delivery vehicle deployment.

Tractable reached a $1B valuation with its AI accident and disaster assessment platform, used by 40+ insurers including GEICO, Ageas, and Tokio Marine to automatically assess vehicle damage from photos. Its CV models process insurance claims in minutes vs. days, reducing settlement time by 10x and cutting fraud by 30%+. In 2025, Tractable expanded from auto insurance into property claims, processing satellite and drone imagery to assess hurricane and wildfire damage across entire zip codes simultaneously.

Scale AI reached a $7.3B valuation as the leading data annotation and AI evaluation platform, labeling 100M+ images and video frames annually for customers including the US Army, OpenAI, Microsoft, and GM. Its Nucleus platform enables teams to find failure modes in CV models and curate the most impactful training data. In 2025, Scale AI was awarded a $249M US Army contract for computer vision data labeling for autonomous vehicle and drone programs.

Landing AI, founded by Andrew Ng, reached an $850M valuation with its LandingLens visual AI platform enabling non-ML engineers to deploy industrial CV inspection without deep learning expertise. Its defect detection systems are deployed in semiconductor fabs, pharmaceutical manufacturing, and food processing, reducing inspection cycle times from hours to milliseconds. In 2025, Landing AI launched Visual Prompting technology that allows CV models to be fine-tuned from just 5-10 labeled examples — 99% less data than traditional approaches.

Cognex generated $800M in 2024 revenue holding 20%+ global machine vision market share with its In-Sight smart cameras and VisionPro software deployed in automotive, electronics, and pharmaceutical manufacturing. Its deep learning ViDi suite detects surface defects invisible to traditional rule-based vision systems. Cognex systems perform over 1 billion quality inspections daily worldwide, and its 2025 Edge Learning launch allows on-device model training on production lines without cloud connectivity.

Matterport generated $39M in Q1 2025 revenue from its 3D spatial intelligence platform with 11M+ scanned spaces spanning real estate, construction, insurance, and facilities management. Its Pro3 camera creates millimeter-accurate digital twins of physical spaces in under 60 minutes. In 2025, CoStar Group acquired Matterport for $1.6B, integrating its spatial scanning technology into the largest commercial real estate data platform and enabling automatic property condition assessment from digital twin data.

Roboflow raised $40M to build the developer platform for computer vision, serving 250,000+ developers and 10,000+ organizations who have trained 100,000+ CV models on the platform. Its dataset management, annotation tools, and model training pipeline compress CV application development from months to days. In 2025, Roboflow launched RF-DETR — its open-source real-time object detection model that outperforms YOLO on COCO benchmarks — becoming the most-starred CV model repository on GitHub within 90 days.

Labelbox reached a $1B valuation with its enterprise data labeling platform processing 100M+ labels per day for Fortune 500 clients including Procter & Gamble, Walmart, and DoorDash. Its Model-Assisted Labeling feature uses active learning to automatically pre-label images, cutting human annotation time by 70%+. In 2025, Labelbox launched Catalog — a semantic search engine for unstructured visual data that lets teams find specific CV training examples using natural language queries across billion-image datasets.

Clarifai raised $100M+ as an enterprise computer vision API platform serving food safety inspection, retail shelf analytics, and defense surveillance use cases. Its multi-modal AI platform processes images, video, text, and audio in a unified pipeline, enabling complex CV workflows like automatic shelf inventory analysis combining product recognition and OCR. In 2025, Clarifai deployed its CV platform with the US Department of Defense for automated aerial surveillance analysis, processing 10,000+ hours of drone footage monthly.

Wayve raised $1.05B in 2024 — the largest single investment in a UK AI company — from SoftBank, Microsoft, and NVIDIA to develop its Embodied AI approach to autonomous driving using imitation learning rather than rule-based systems. Its AV2.0 model learns to drive by watching humans, then generalizes to new roads without explicit programming. In 2025, Wayve launched US commercial pilots in San Francisco and Austin, and its foundation model technology was licensed to Amazon for autonomous delivery vehicle deployment.

Tractable reached a $1B valuation with its AI accident and disaster assessment platform, used by 40+ insurers including GEICO, Ageas, and Tokio Marine to automatically assess vehicle damage from photos. Its CV models process insurance claims in minutes vs. days, reducing settlement time by 10x and cutting fraud by 30%+. In 2025, Tractable expanded from auto insurance into property claims, processing satellite and drone imagery to assess hurricane and wildfire damage across entire zip codes simultaneously.