# Datature

Datature is an end-to-end Vision AI platform for building, annotating, training, and deploying computer vision models. It replaces the typical multi-tool stack with a single platform covering the full pipeline from raw images and video to production inference on cloud, edge, and on-premises environments.

Datature serves teams that need production-grade computer vision without dedicated ML infrastructure staff. The platform supports image classification, object detection, keypoint detection, semantic and instance segmentation, and video classification. It includes 50+ model architectures (YOLO variants, EfficientNet, Mask R-CNN, transformer-based models), AI-assisted annotation tools, pre-configured training pipelines, and deployment to edge hardware (NVIDIA Jetson, MemryX, Hailo). Datature is one of the few platforms with native support for 3D medical imaging formats (NIfTI/DICOM) with multi-planar annotation.

Datature Vi is a new product that allows teams to fine-tune large vision-language models (VLMs), particularly CosmosReason2, Qwen, InternVL, and more. This product support users who wish to build their own VLM on their own data, and leveraging Datature's highly-efficient data pipelines. Datature is the only platform that supports both geometric models and VLMs end-to-end.

Over 10,000 teams across manufacturing, healthcare, agriculture, energy, logistics, sports, and defense in 53+ countries use Datature. Notable customers include FedEx, Toyota, NHS, Wien Energie, Rolls-Royce, and SATS.

## Product

- [Datature Label](https://datature.io/product/label): AI-assisted annotation tools supporting bounding boxes, polygons, segmentation masks, keypoints, and classification across images, video, and 3D medical formats
- [Datature Train](https://datature.io/product/train): Drag-and-drop workflow builder for model training with 50+ architectures, hyperparameter tuning, augmentation, and real-time training dashboards
- [Datature Deploy](https://datature.io/product/deploy): Model deployment to cloud APIs, edge devices (Jetson, MemryX, Hailo), and on-premises infrastructure with optimization and monitoring
- [Datature Services](https://datature.io/product/services): Professional services for enterprise implementation, custom development, and dedicated support

## Documentation

- [Introduction to Datature](https://developers.datature.io/docs/introduction-to-datature): Platform overview and core concepts
- [Quickstart Guide](https://developers.datature.io/docs/quickstart): Step-by-step setup for new users
- [Model Training Tutorials](https://developers.datature.io/docs/model-training-tutorials): Guides for training models across task types
- [Annotation Tools](https://developers.datature.io/docs/annotations): Documentation for all annotation features and workflows
- [Workflow Builder](https://developers.datature.io/docs/workflows): Creating and configuring training pipelines
- [Deployment Guide](https://developers.datature.io/docs/making-predictions): Deploying models and running inference
- [Outpost (Edge Deployment)](https://developers.datature.io/docs/outpost): Deploying and managing models on edge devices
- [API Deployments](https://developers.datature.io/docs/api-deployments): Cloud-hosted model inference via REST API

## Developer Tools

- [Python SDK](https://developers.datature.io/docs/python-sdk): Programmatic access to workspaces, projects, assets, annotations, training, and deployment
- [CLI Tool](https://developers.datature.io/docs/cli): Command-line interface for project and asset management
- [REST API Reference](https://developers.datature.io/reference): Complete API endpoint documentation for all platform operations
- [SDK Changelog](https://developers.datature.io/docs/sdk-changelog): Version history and updates for the Python SDK

## Tutorials

- [Build an Object Detection Model](https://datature.io/tutorials/build-an-object-detection-model): End-to-end tutorial for object detection
- [Build an Image Classification Model](https://datature.io/tutorials/build-an-image-classification-model): End-to-end tutorial for image classification
- [Build a Semantic Segmentation Model](https://datature.io/tutorials/built-a-semantic-segmentation-model): End-to-end tutorial for semantic segmentation
- [Build a Keypoint Estimation Model](https://datature.io/tutorials/build-a-keypoint-estimation-model): End-to-end tutorial for keypoint and pose estimation
- [Build a Video Classification Model](https://datature.io/tutorials/build-a-video-classification-model): End-to-end tutorial for video classification

## Solutions

- [Healthcare](https://datature.io/solutions/healthcare): Pathology slide analysis, radiology screening, 3D brain imaging with NIfTI/DICOM support
- [Manufacturing](https://datature.io/solutions/manufacturing): Defect detection, dimensional measurement, visual quality control automation
- [Agriculture](https://datature.io/solutions/agriculture): Crop health monitoring, pest detection, yield measurement
- [Retail](https://datature.io/solutions/retail): Inventory tracking, shelf analysis, customer behavior analytics
- [Smart City](https://datature.io/solutions/smart-city): Traffic monitoring, infrastructure inspection, urban planning
- [Utilities](https://datature.io/solutions/utilities): Equipment monitoring, infrastructure inspection, anomaly detection

## Company

- [About Datature](https://datature.io/company): Company background, team, and mission
- [Pricing](https://datature.io/pricing): Free, Developer, Professional, and Enterprise plans with details on assets, storage, training compute, and deployment limits
- [Security](https://datature.io/security): SOC 2 Type 2, HIPAA compliance, encryption at rest and in transit, penetration testing, and data isolation practices
- [Privacy Policy](https://datature.io/privacy): Data handling and privacy practices
- [Terms of Service](https://datature.io/terms-of-service): Platform terms and conditions
- [Legal](https://datature.io/legal): Legal hub and compliance documentation

## Contact

- [Contact Datature](https://datature.io/contact): General inquiries and support
- [Contact Sales](https://datature.io/contact/sales): Enterprise pilots, pricing, and demos
- [Contact Support](https://datature.io/contact/support): Technical support and troubleshooting

## Resources

- [Blog](https://datature.io/blog): Technical guides, product announcements, industry insights, and customer case studies
- [Glossary](https://datature.io/glossary): Definitions of computer vision and machine learning terminology
- [Events](https://datature.io/events): Webinars, product launches, and community events
- [Developer Documentation](https://developers.datature.io): Full platform documentation and API reference
- [Platform Changelog](https://developers.datature.io/changelog): Release notes and platform updates
- [GitHub](https://github.com/datature): Open-source scripts, notebooks, and example workflows
- [Datature Nexus App](https://nexus.datature.io): Sign in to the Datature platform

## Blog

- [VLM Training Metrics and Loss Functions: A Technical Reference [2026]](https://datature.io/blog/vlm-evaluation-metrics-guide-2026)
- [An Introduction to BYTETrack: Multi-Object Tracking by Associating Every Detection Box](https://datature.io/blog/introduction-to-bytetrack-multi-object-tracking-by-associating-every-detection-box)
- [The Enterprise Vision AI Adoption Report 2026](https://datature.io/blog/enterprise-vision-ai-adoption-report-2026)
- [Finetuning Your Own Cosmos-Reason2 Model](https://datature.io/blog/finetuning-your-own-cosmos-reason2-model)
- [Visual Question Answering: A Comprehensive Guide to Fine-tuning VLMs for Intelligent Image Understanding](https://datature.io/blog/visual-question-answering-a-comprehensive-guide-to-fine-tuning-vlms-for-intelligent-image-understanding)
- [Introduction to Chain-of-Thought for Vision-Language Models](https://datature.io/blog/introduction-to-chain-of-thought-for-vision-language-models)
- [Containerized VLM Deployment: A Practical Guide to NVIDIA NIM](https://datature.io/blog/containerized-vlm-deployment-a-practical-guide-to-nvidia-nim)
- [A Comprehensive Guide to Object Tracking Algorithms in 2025](https://datature.io/blog/a-comprehensive-guide-to-object-tracking-algorithms-in-2025)
- [SAM 3: A Technical Deep Dive into Meta's Next-Generation Segmentation Model](https://datature.io/blog/sam-3-a-technical-deep-dive-into-metas-next-generation-segmentation-model)
- [Introducing t-SNE Embedding Visualization on Datature: Discover Image Similarity and Patterns](https://datature.io/blog/introducing-t-sne-embedding-visualization-on-datature-discover-image-similarity-and-patterns)
- [How to Fine-Tune Qwen2.5-VL](https://datature.io/blog/how-to-fine-tune-qwen2-5-vl)
- [A Comprehensive Guide to Model Fusion Techniques for Metadata-Aware Training](https://datature.io/blog/a-comprehensive-guide-to-model-fusion-techniques-for-metadata-aware-training)
- [Solving Class Imbalance: Upsampling in Machine Learning Projects](https://datature.io/blog/solving-class-imbalance-upsampling-in-machine-learning-projects)
- [A Comprehensive Guide to 3D Models for Medical Image Segmentation](https://datature.io/blog/a-comprehensive-guide-to-3d-models-for-medical-image-segmentation)
- [Introducing PaliGemma 2: Use Cases and Improvements](https://datature.io/blog/introducing-paligemma-2-use-cases-and-improvements)
- [Beyond SAM-2: Exploring Derivatives for Better Performance](https://datature.io/blog/beyond-sam-2-exploring-derivatives-for-better-performance)
- [SAM2Long: Higher Precision in Long-Term Video Segmentation](https://datature.io/blog/sam2long-higher-precision-long-term-video-segmentation)
- [Real-Time Object Detection With D-FINE](https://datature.io/blog/real-time-object-detection-d-fine)
- [An Introduction to Bitmask Representations and Encodings - RLE vs REE](https://datature.io/blog/an-introduction-to-bitmask-representations-and-encodings-rle-vs-ree)
- [How to Use LiteRT for Real-Time Inferencing on Android](https://datature.io/blog/how-to-use-litert-for-real-time-inferencing-on-android)
- [YOLO11: Step-by-Step Training on Custom Data and Comparison with YOLOv8](https://datature.io/blog/yolo11-step-by-step-training-on-custom-data-and-comparison-with-yolov8)
- [Introducing Class Metrics and Low Confidence Sampling for Deeper Model Evaluation Insights](https://datature.io/blog/introducing-class-metrics-and-low-confidence-sampling-for-deeper-model-evaluation-insights)
- [Introducing Meta Segment Anything Model 2: Use Cases and Improvements](https://datature.io/blog/introducing-metas-sam-2-use-cases-and-improvements)
- [Annotate Images & Videos with Segment Anything 2.0 on Datature Nexus](https://datature.io/blog/annotate-images-videos-with-segment-anything-2-0-on-datature-nexus)
- [Introducing Florence-2: Microsoft's Latest Multi-Modal, Compact Visual Language Model](https://datature.io/blog/introducing-florence-2-microsofts-latest-multi-modal-compact-visual-language-model)
- [A Primer on Fine-Tuning PaliGemma and VLMs](https://datature.io/blog/a-primer-on-fine-tuning-paligemma-and-vlms)
- [Introducing PaliGemma: Google's Latest Visual Language Model](https://datature.io/blog/introducing-paligemma-googles-latest-visual-language-model)
- [Introducing Vision Transformers for Robust Segmentation](https://datature.io/blog/introducing-vision-transformers-for-robust-segmentation)
- [Introducing MoViNet for Video Classification](https://datature.io/blog/introducing-movinet-for-video-classification)
- [Ingroth Automates Barrel Defect Inspection with Datature](https://datature.io/blog/ingroth-automates-barrel-defect-inspection-with-datature)
- [Trendspek Optimizes Structural Crack Detection with Datature](https://datature.io/blog/trendspek-optimizes-structural-crack-detection-with-datature)
- [A Comprehensive Guide to Neural Network Model Pruning](https://datature.io/blog/a-comprehensive-guide-to-neural-network-model-pruning)
- [YOLOv9 - A Comprehensive Guide and Custom Dataset Fine-Tuning](https://datature.io/blog/yolov9-a-comprehensive-guide-and-custom-dataset-fine-tuning)
- [Introducing Ontologies for Advanced Annotation](https://datature.io/blog/introducing-ontologies-for-advanced-annotation)
- [Introducing Post-Training Model Quantization Feature and Mechanics Explained](https://datature.io/blog/introducing-post-training-quantization-feature-and-mechanics-explained)
- [Introducing SAHI and Sliding Window Functions for Small Object Detections](https://datature.io/blog/sahi-and-sliding-window-functions-for-small-object-detections)
- [How to Perform Action Recognition on Keypoints with ST-GCN++](https://datature.io/blog/how-to-perform-action-recogition-on-keypoints-with-stgcnn)
- [Training Image Classification Models with YOLOv8 on Nexus](https://datature.io/blog/training-image-classification-models-with-yolov8-on-nexus)
- [Understanding Your YOLOv8 model with Eigen-CAM](https://datature.io/blog/understanding-your-yolov8-model-with-eigen-cam)
- [Introducing Advanced Search for Exploring and Managing Data](https://datature.io/blog/introducing-advanced-search-for-exploring-and-managing-data)
- [Leveraging Datature Nexus for Tumor and Anomaly Detection in Medical Scans : Part I](https://datature.io/blog/leveraging-datature-nexus-for-tumor-and-anomaly-detection-in-medical-scans)
- [How to Evaluate Computer Vision Models with Confusion Matrix](https://datature.io/blog/how-to-evaluate-computer-vision-models-with-confusion-matrix)
- [Label Your Data with Segment Anything on Nexus](https://datature.io/blog/label-your-data-with-segment-anything-on-nexus)
- [How to Build Your Own AI-Generated Image with ControlNet and Stable Diffusion](https://datature.io/blog/how-to-build-your-own-ai-generated-image-with-controlnet-and-stable-diffusion)
- [Leveraging Active Learning to Optimize Your Computer Vision Pipeline](https://datature.io/blog/leveraging-active-learning-to-optimize-your-computer-vision-pipeline)
- [A Guide to Using DeepLabV3 for Semantic Segmentation](https://datature.io/blog/a-guide-to-using-deeplabv3-for-semantic-segmentation)
- [How to Assess Your Labelling Metrics with Performance Tracking](https://datature.io/blog/how-to-assess-your-labelling-metrics-with-performance-tracking)
- [How to Improve Data Annotation Quality with Consensus Algorithm](https://datature.io/blog/how-to-improve-data-annotation-quality-with-consensus-algorithm)
- [How to Connect Azure Blob Storage for Asset Uploading](https://datature.io/blog/how-to-connect-azure-blob-storage-for-asset-uploading)
- [Accelerating Video Annotation with Video Interpolation/Video Tracking](https://datature.io/blog/accelerating-video-annotation-with-video-interpolation-video-tracking)
- [How to Automate Your MLOps Pipeline with Python SDK](https://datature.io/blog/how-to-automate-your-mlops-pipeline-with-python-sdk)
- [How to Use Annotation Automation to Organize Your Labelling Workforce](https://datature.io/blog/how-to-use-annotation-automation-to-organize-your-labelling-workflow)
- [How to Annotate Videos Data on Datature Nexus](https://datature.io/blog/how-to-annotate-videos-data-on-datature-nexus)
- [A Historical Breakdown of YOLO: A Landmark Model in Object Detection](https://datature.io/blog/a-historical-breakdown-of-yolo)
- [Supporting Fully Convolutional Networks (and U-Net) for Image Segmentation](https://datature.io/blog/supporting-fully-convolutional-networks-and-u-net-for-image-segmentation)
- [Gardyn Optimizes Plants' Growth With Datature](https://datature.io/blog/gardyn-optimizes-plants-growth-with-datature)
- [Get Started with Training a YOLOv8 Object Detection Model](https://datature.io/blog/get-started-with-training-a-yolov8-object-detection-model)
- [How To Interpret Training Graphs to Understand and Improve Model Performance](https://datature.io/blog/how-to-interpret-training-graphs-to-understand-and-improve-model-performance)
- [MacroInsight Builds Clinical Decision Support Systems with Datature](https://datature.io/blog/macroinsight-builds-clinical-decision-support-systems-with-datature)
- [Introducing Datature Python SDK and Management REST API](https://datature.io/blog/introducing-datature-python-sdk-and-management-rest-api)
- [How to Use the External Labelling Service on Nexus](https://datature.io/blog/how-to-use-the-external-labelling-service-on-nexus)
- [Introducing Our External Labelling Service to Accelerate Your Ground Truth](https://datature.io/blog/introducing-our-external-labelling-service-to-accelerate-your-ground-truth)
- [Introduction to Multiple Object Tracking and Recent Developments](https://datature.io/blog/introduction-to-multiple-object-tracking-and-recent-developments)
- [Introducing Video Compatibility to Support Broader Use Cases in Computer Vision](https://datature.io/blog/introducing-video-compatibility-to-support-broader-use-cases-in-computer-vision)
- [How to Connect Amazon S3 Bucket for Asset Uploading](https://datature.io/blog/how-to-connect-amazon-s3-bucket-for-asset-uploading)
- [Introducing Model Assisted Labelling to Streamline Your MLOps Pipeline](https://datature.io/blog/introducing-model-assisted-labelling-to-streamline-your-mlops-pipeline)
- [CentrovisioN Improves Building Inspection Processes with Datature](https://datature.io/blog/how-centrovision-uses-datature-platform-to-improve-building-inspection-processes)
- [Introducing Asset Group Management for Managing Datasets](https://datature.io/blog/introducing-asset-group-management-for-managing-datasets)
- [Exploring Real-World Use Cases with Datature Nexus](https://datature.io/blog/exploring-real-world-use-cases-with-datature-nexus)
- [Building a Simple Inference Dashboard with Streamlit](https://datature.io/blog/building-a-simple-inference-dashboard-with-streamlit)
- [How to Load Vision Models on Raspberry Pi for Edge Deployment](https://datature.io/blog/how-to-load-vision-models-on-raspberry-pi-for-edge-deployment)
- [Learn How to Export TFLite Models for Object Detection Models](https://datature.io/blog/how-to-export-tflite-models-for-object-detection-models)
- [Running Custom TensorFlow.js Object Detection Models in Node-RED](https://datature.io/blog/running-custom-tensorflow-js-object-detection-models-in-node-red)
- [How to Visually Inspect Your Dataset and Annotations for Model Training](https://datature.io/blog/visually-inspect-your-dataset-and-annotations-for-model-training)
- [How To Train YOLOX Object Detection Model On A Custom Dataset](https://datature.io/blog/train-yolox-object-detection-model-on-a-custom-dataset)
- [Introducing AI Mask Refinement To Enhance Annotations](https://datature.io/blog/introducing-ai-mask-refinement-to-enhance-annotations)
- [Performing Image Augmentation For Machine Learning](https://datature.io/blog/performing-image-augmentation-for-machine-learning)
- [How To Use API Deployment For Trained Model Inference](https://datature.io/blog/how-to-use-api-deployment-for-trained-model-inference)
- [Comprehensive Guide to Zero-Shot and K-Shot Learning](https://datature.io/blog/zero-shot-and-k-shot-learning)
- [BrainScanology Optimizes Medical AI Models with Datature](https://datature.io/blog/how-brainscanology-uses-datatures-platform-to-build-medical-ai-models-that-revolutionize-image-based-diagnostics)
- [Implementing Object Tracking for Computer Vision (+ Code)](https://datature.io/blog/implementing-object-tracking-for-computer-vision)
- [Announcing Our Series Seed Funding](https://datature.io/blog/announcing-our-series-seed-funding)
- [Learning ASL with Computer Vision](https://datature.io/blog/learning-asl-with-computer-vision)
- [Improving Datasets & ML Models with Metadata Query](https://datature.io/blog/datature-metadata-query)
- [Train and Visualize A Face Mask Detection Model Without Code](https://datature.io/blog/train-and-visualize-face-mask-detection-model)
- [Using AI-Guided Segmentation To Speed Up Labelling With IntelliBrush](https://datature.io/blog/using-ai-guided-segmentation-to-speed-up-labelling-with-intellibrush)
- [Inspect Model Inferences on Images and Videos with Portal](https://datature.io/blog/inspect-model-inferences-on-images-and-videos-with-portal)
- [Training an Instance Segmentation Model with Custom Data](https://datature.io/blog/training-an-instance-segmentation-model)
- [The Five Minutes Datature Demo](https://datature.io/blog/five-minutes-datature-demo)