Harvey launches Tenet legal-specialized model post-trained on Moonshot Kimi K3
US legal AI unicorn Harvey launched Tenet on Aug 23, a legal-specialized model post-trained on Kimi K3, improving LAB all-pass rates by 82%.
On August 23, US legal AI company Harvey launched Harvey II and a specialized legal model called Tenet. Tenet is post-trained on Moonshot AI's open-weight Kimi K3 model, with inference infrastructure provided by Fireworks AI.
Harvey reports that Tenet improves all-pass rates by 82% on LAB legal benchmarks while costing less than one-fourth of leading foundation models. The release includes three specialized subagent models for M&A diligence, document review, and firm knowledge management.
Harvey previously relied primarily on OpenAI and Anthropic closed-source models for its legal AI services, serving over 2,400 institutions and more than 200,000 lawyers. Tenet marks the first time Harvey has integrated a Chinese open-weight model into its training pipeline. Harvey is currently valued at $11 billion with annualized revenue of approximately $350 million, and is in the midst of raising a new round at around $15.5 billion.