Training data

49 vendors.

• Abundant — Builds RL environments and datasets for AI labs.
• AfterQuery — Records how working professionals reason through real tasks and packages it as training data for frontier model developers.
• Anthromind — Human evaluation, RLHF preference data and post-training datasets with a scalable-oversight focus.
• Appen — Publicly listed veteran of the data industry offering training data, annotation and model evaluation with a large global crowd.
• Bespoke Labs — Data-curation research shop that builds environments and optimizes agents for enterprises and labs.
• Build AI — Positions itself as a large-scale data provider for physical-AI research teams.
• Centific — Provides multilingual datasets, secure data operations and responsible-AI services to model labs and enterprises.
• Chakra — Research and design lab producing expert-authored task sets for training and benchmarks for evaluation.
• Claru — Training-data company focused entirely on physical AI, with a large library of egocentric video and annotations.
• Collinear — Simulation lab producing training data and RL environments so agents can be tested in realistic scenarios.
• Datacurve — Produces custom data for long-horizon reasoning, software engineering and data-science tasks for frontier labs.
• DataOcean AI — Data provider offering off-the-shelf datasets and custom collection and labeling, especially speech and multilingual.
• Datoric — Robotics training data—egocentric video, manipulation and custom teleoperation—with documented provenance.
• David AI — Audio-data research company building proprietary speech and conversation datasets for AI labs.
• Deccan AI — Supplies post-training data and RL environments for coding, multimodal and physical AI, plus off-the-shelf datasets and enterprise agent services.
• Defined.ai — Marketplace and services for multilingual, multimodal training datasets with compliance focus.
• Encord — Multimodal data management and annotation platform with an emphasis on physical-AI and sensor/video data.
• Gradient Health — De-identified multimodal patient data for medical-AI developers through its Atlas platform.
• Handshake AI — The campus-jobs network runs a paid fellowship that recruits experts and students to produce AI training and evaluation work.
• iMerit — Expert-led annotation and fine-tuning services backed by its own workflow tooling.
• Innodata — Listed data-engineering firm providing training data, annotation, evaluation and red-teaming.
• Invisible Technologies — Mixes an expert workforce with orchestration software; supplies post-training data to labs and production AI workflows to enterprises.
• Kled — Licenses curated datasets (video, music, transcripts with metadata) sourced from rightsholders for model training.
• Labelbox — Started as labeling software; now sells expert data and training environments to frontier labs, plus an agent product for engineering teams.
• Lightwheel — Makes simulation-ready 3D assets, synthetic and egocentric human data, and evaluation platforms for embodied AI.
• Luel — Multimodal data lab working on machine perception, supplying rights-cleared contributor data.
• LXT — Collects, annotates and evaluates audio, speech, text, image and video data across a very wide range of language locales.
• Mecka — Captures and structures real-world human activity into datasets and evaluations for robotics teams.
• Mercor — Recruits domain experts at scale and turns their work into training data, benchmarks and agent-training tasks for labs and enterprises.
• Midcentury — First-person video of real work plus high-fidelity simulation for robot training.
• Pareto AI — Expert human-data provider focused on feedback signals that improve frontier models.
• Patronus AI — Research lab building training data and simulated environments for frontier AI, evolving from an evaluation-tools origin.
• Proximal — Research lab producing coding data for autonomous coding agents.
• Revelo — Latin American engineering-talent company that now sells code-focused expert data, RL environments and evals to AI labs.
• Sama — Annotation company covering computer vision and generative-AI data with human-in-the-loop quality control.
• Scale AI — Long-running supplier of labeled data, human feedback and model evaluations to frontier labs, government agencies and large enterprises.
• Sieve — Multimodal data lab supplying video, audio, image and interaction data to frontier AI.
• Snorkel AI — Builds specialized datasets, benchmarks and evaluation environments aimed at high-stakes domains for frontier models and agents.
• Surge AI — Human-data company focused on high-skill annotators and RLHF data for frontier model training.
• Symbal — Trains models and deploys them for enterprises while supplying post-training data to frontier labs.
• Taste Labs — Builds infrastructure for bringing brand and creative IP into AI systems, including design-preference data.
• TELUS Digital — Digital-services arm of TELUS offering AI data collection, annotation and evaluation at scale.
• Toloka — Data partner for agentic, coding and safety work, blending human experts with tooling to generate training and evaluation data.
• Troveo — Sources permissioned, non-public business data plus large video, audio and gaming libraries for frontier labs.
• TrustScale (formerly e2f) — Long-time language-data provider, now branded TrustScale, that also markets a hallucination-verification product.
• Turing — Draws on a large engineering talent network to produce coding and reasoning data and help enterprises deploy agents.
• Vetto — Post-training partner covering RL environments, rubrics and verifiers, RLHF, SFT and expert human evaluation.
• Wirestock — Pays a large creator community to produce photos, video and design work used as licensed AI training data.
• XDOF — Robotics company collecting teleoperation data and motion capabilities for autonomous systems.