Agentic Annotation Platform
Open Source

CaliperGT

Transform raw sensor data into trustworthy ground truth — faster, with quality you can measure.

See pricing & editions
CaliperGT point cloud annotation

Platform Features

Comprehensive tools for every stage of your annotation workflow

Intelligent Agentic Annotation

  • AI-powered annotation for LiDAR point clouds, images, and radar data
  • Pre-trained models for common object classes with custom model support
  • Batch processing capabilities for large datasets
Open Source

Open source. Yours to run.

Deploy the complete CaliperGT annotation platform on your own infrastructure — free, with no license fee. Keep your data in-house and extend it to your workflow.

bash
# deploy CaliperGT on your own infrastructure
$ git clone https://github.com/caliperai-ai/caliperai-gt
$ cd caliperai-gt
# follow the README to launch
caliperai-ai/caliperai-gt

Star it, fork it, open an issue — build with us in the open.

View on GitHub

Self-hosted

Run it in your own cloud, data center, or dev environment.

SSO/OIDC + RBAC

Enterprise-grade access control, included out of the box.

Your data stays in-house

Sensor data never has to leave your infrastructure.

Full platform

Image, video & LiDAR annotation with AI assist — no feature gates.

Proven Results

Real performance improvements from teams using CaliperGT

Up to 5× faster

dataset creation

Up to 40% reduction

in manual effort

Improved consistency

via guided QA workflows

* Outcomes vary by dataset and process. Results based on customer case studies.

Great ground truth starts with great data

Log synchronized camera, LiDAR, and sensor data with CaliperCapture, then turn it into training-ready ground truth in CaliperGT.

Explore CaliperCapture

Workflow Timeline

From raw data to production-ready annotations in six streamlined steps

Ingest

Upload multimodal sensor data

Auto-Annotate

AI-powered initial annotations

Review/QA

Human validation and corrections

RLHF Feedback

Model learns from corrections

Export

Standard format outputs

Integrate

MLOps pipeline integration

Integrations

Export to standard formats and integrate with your existing ML pipeline

Export Formats
COCO
KITTI
NuScenes
Datumaro
YOLO
Pascal VOC
Cityscapes
Open3D
APIs & SDKs
REST API
GraphQL
Python SDK
CLI Tools
Webhooks
Batch Export
Platform Integrations
AWS S3
Google Cloud
Azure Blob
MLflow
Weights & Biases
DVC
Annotation Services

Professional Annotation Services

Leverage our expert team and strategic partnerships to deliver high-quality annotation services for your autonomous driving and computer vision projects.

Annotation Speed: 1000+ 3D Bounding Boxes per Day per Person
LiDAR Point Cloud Annotation
Precise 3D bounding boxes, semantic segmentation, and instance segmentation for autonomous driving datasets
  • 3D bounding boxes
  • Semantic segmentation
  • Instance segmentation
  • Trajectory annotation
Image & Video Annotation
Comprehensive 2D annotation services for computer vision applications
  • Object detection
  • Semantic segmentation
  • Instance segmentation
  • Video tracking
Sensor Fusion Annotation
Comprehensive annotation for multi-sensor fusion data in autonomous systems
  • Multi-sensor fusion
  • Temporal tracking
  • Cross-modal alignment
  • Fusion validation
Our Expert Team
A dedicated team of annotation specialists with deep domain expertise
  • Expert annotators with 5+ years experience
  • Domain specialists in autonomous driving
  • Quality assurance specialists
  • Project managers with technical backgrounds
  • 24/7 support and turnaround times
Strategic Partnerships
Leveraging partnerships with leading annotation companies for scale and expertise
  • Strategic partnerships with leading annotation companies
  • Access to global annotation workforce
  • Scalable capacity for large projects
  • Diverse expertise across multiple domains
  • Cost-effective solutions through economies of scale

Ready to Scale Your Annotation Projects?

Get in touch with our team to discuss your annotation requirements and learn how we can help accelerate your AI development with high-quality, scalable annotation services.

Simple editions, serious capability

Start free with the self-hosted open-source platform, or upgrade to Enterprise for automation, scale, security, and managed operations.

Community Edition

Open Source

Free

Self-hosted. No platform license fee.

For engineering teams, researchers, startups, and internal AI teams that want to manage and annotate automotive multimodal data on their own infrastructure.

  • Automotive multimodal data managementOrganize campaigns, datasets, scenes, sensor streams, annotations, and metadata.
  • High-performance LiDAR visualizationOptimized point-cloud loading and rendering for responsive 3D workflows.
  • Image, video and LiDAR annotationCore 2D and 3D labeling workflows in one platform.
  • AI-assisted annotation toolsAssisted tracking, segmentation, lanes, and productivity tools.
  • Team access and workflow controlsSSO/OIDC integration, role-based access control, and project permissions.
  • Self-hosted deploymentRun the platform within your cloud, data center, or development environment.

Community support through documentation, GitHub issues, and product discussions.

Recommended for production

Enterprise

Custom pricing

Based on users, deployment, data volume, automation, and support requirements.

Software subscriptions start from $19 / user / month. Infrastructure, model inference, storage, managed services, and usage-based automation are priced separately.

For automotive, ADAS, robotics, mapping, and autonomous-system teams that need automation, scale, accuracy, governance, and enterprise support.

  • VLA (vision-language-action) labellingLanguage- and action-grounded ground truth for vision-language-action models across driving and robotics scenes.
  • 3D LiDAR auto-labelingModel-in-the-loop 3D box detection, multi-object tracking, static-object aggregation, and 4D long-sequence labeling.
  • 2D auto-labeling and trackingHigh-accuracy detection, segmentation, and tracking across camera streams with model-in-the-loop pre-labeling.
  • Calibration-aware data workflowsCamera, LiDAR, ego-pose, timestamp, projection, and sensor-alignment validation.
  • Automotive multimodal data ingestionStructured ingestion of camera, LiDAR, CAN, GNSS, IMU, ROS, ROS 2, and MCAP sensor data into the platform.
  • Foundation-model-assisted data improvementUse CaliperAI fine-tuned foundation models to improve pre-label quality, surface difficult samples, and accelerate dataset refinement.
  • Enterprise security, governance and supportAdvanced access controls, auditability, private deployment, onboarding, integrations, and SLAs.

Available as self-hosted, private cloud, on-premise, or managed deployment.

Full pricing details are available at /pricing.

Ready to transform your annotation workflow?

See CaliperGT in action with a personalized demo tailored to your use case and data requirements.

Request a CaliperGT Demo