Veritone, Inc. (VERI) announced a major artificial intelligence milestone, revealing that its Veritone Data Refinery processed more than 5 trillion tokens in the second quarter of 2025. The achievement underscores Veritone’s growing role in delivering high-quality, multimodal training data for enterprise AI solutions.
The VDR platform, powered by Veritone’s aiWARE operating system, ingests and structured unstructured media such as video archives, audio libraries, security footage, and marketing content and converts them into AI-ready tokens with embedded governance and rights metadata. These structured datasets are fueling new AI models while creating licensing and monetization opportunities for content owners across media, entertainment, and the public sector.
Management highlighted that VDR has already surpassed a $20 million near-term sales pipeline, up 33% from June 2025 estimates and 100% from Q1 levels, signaling accelerating demand for curated, compliant AI data.
Veritone emphasized its leadership in four core areas: multimodal AI data processing, unlocking unstructured data, solving complex enterprise challenges, and enabling new revenue streams for rights holders.
The company noted that high-quality training data is increasingly recognized as the foundation of next-generation AI, and Veritone’s tokenization at scale puts it in a strong position to capture market share as enterprises seek reliable and compliant data sources.
The post Veritone Hits 5 Trillion Token Milestone, Expands AI Training Data Leadership appeared first on PRISM MarketView.
