Why are there six different marketplace pages?

Each marketplace optimises for a different buyer intent. A buyer asking for VLA training data has different requirements than a buyer asking for evaluation data, even when the underlying capture partner overlaps. Routing each intent to its own intake page keeps the spec accurate.

Can a buyer combine intents across marketplaces?

Yes. Most large physical AI programs need both training and evaluation data, often across modalities. Use the cross-marketplace decision table on this page to identify which intake page to start from, then add the others as the program scope expands.

Which marketplace handles deployment-environment data?

Deployment-specific data is captured against the buyer's robot, environment, and task — typically routed through /robot-training-data-marketplace or /teleoperation-data-marketplace depending on whether the policy is action-producing.

Buyer routing

Physical AI data marketplaces — when to use which

Six intake pages cover the most common physical AI buyer intents. The right one depends on modality (egocentric vs teleop vs robot-demos), output (VLA pair vs eval set), and scale (full capture vs validation sample). Use the decision table to route to the marketplace that matches your spec.

Decision rule

Pick the marketplace that matches your intent
If you need…	Start at	Why
I need broad physical AI data — discovery and selection	/physical-ai-data-marketplace	Cross-modality marketplace covering teleop, egocentric, demonstrations, and eval.
First-person / wearable / hand-pose / consented egocentric	/egocentric-data-licensing	Egocentric capture with explicit licensing review and consent artifacts.
Robot teleoperation traces with synchronized state and action	/teleoperation-data-marketplace	Teleop-specific intake — robot embodiment, action stream, success/failure flags.
Robot demonstrations and trajectories for policy training	/robot-training-data-marketplace	Generalist robotics intake covering manipulation, navigation, eval sets.
Vision-language-action paired data for OpenVLA / RT-2 / π0 / GR00T	/vla-training-data	VLA-specific routing — paired observations + instructions + actions.
Smaller eval / validation / benchmark sample before scale	/eval-data-for-robotics	Eval-bound intake with scoped sample size and acceptance criteria.

All 7 marketplaces

Egocentric video data hub

Egocentric Video Data: Capture, License & Deliver for Physical AI

Egocentric video data is first-person video recorded from a head-mounted or wearable camera while a person performs a real-world task. It teaches physical-AI and robotics models hand-object interaction and viewpoint grounding. Truelabel sources it from 20,000+ consented collectors worldwide and delivers it robotics-ready, with full provenance.

Buyer intake

Physical AI data marketplace

Truelabel is a physical AI data marketplace where 100+ vetted capture partners deliver egocentric video, teleoperation traces, robot demonstrations, and evaluation datasets with commercial-use rights, contributor consent artifacts, and per-buyer fitness review. US search demand for 'physical AI' grew 3.5× between May 2025 and April 2026 (1,900 → 6,600 monthly searches), driven by humanoid programs (NVIDIA GR00T N1, Figure AI commercial deployments), open-source policy releases (Open X-Embodiment, RoboCasa), and the maturation of factory-style data pipelines that replace passive web scraping. Buyers post a sourcing request, review matched samples, and ingest data with rights and metadata attached.

First-person data

Egocentric data licensing

Egocentric data is first-person video or sensor data captured from the perspective of a person performing real tasks — truelabel sources licensed egocentric footage benchmarked against public corpora like Ego4D's 3,670 hours of first-person daily-life activity. Buyers define environment, task, consent, and metadata requirements before suppliers submit samples for review.

Robot action data

Teleoperation data marketplace

A teleoperation data marketplace lets robotics teams source synchronized camera streams, joint states, end-effector poses, action traces, task labels, and success/failure metadata captured while a human operator controls the robot — and Truelabel benchmarks teleop sourcing against public references like RoboSet's 9,500 teleoperated trajectories. Truelabel matches teleop sourcing requests to vetted capture partners and routes samples through buyer review before scale.

Robotics datasets

Robot training data marketplace

A robot training data marketplace coordinates demonstrations, trajectories, video, robot state, action streams, and evaluation sets across vetted capture partners. Truelabel converts buyer requirements into supplier-facing specs, routes them to capture partners with the right embodiment and rights posture, and runs sample review before scale.

Vision-language-action models

VLA training data

VLA training data is the paired vision, language, and action signal that vision-language-action models — OpenVLA, RT-2, π0, GR00T N1 — train on: robot teleoperation traces, instruction-grounded demonstrations, egocentric video, action labels, and timing metadata. Truelabel routes VLA sourcing requests to capture partners that can deliver paired modalities at the buyer's target embodiment.

Fast validation

Eval data for robotics

Robotics eval data is a smaller dataset used to test model behavior, supplier quality, or task coverage before a larger training-data buy. truelabel's eval-request path lets buyers source a pre-scoped sample set with rights, consent, metadata, and acceptance criteria attached.