Harvey AI Corp. has raised $550 million in a funding round that values the legal technology company at $15.5 billion, just six months after its previous nine-figure investment. Diffusion and Lightspeed Venture Partners led the deal, with participation from more than a dozen other backers, including Sequoia, Kleiner Perkins and Goldman Sachs. The company sells a cloud platform that law firms use to automate manual attorney work, and it reports that 80% of the 100 highest-ranked U.S. law firms are among its installed base. In-house legal teams at large enterprises also make up a significant customer segment, with half of the Fortune 10 using the platform.
The funding arrives days after Harvey introduced its first custom large language model, called Tenet. Tenet is a fine-tuned version of Kimi K3, an open-source LLM with 2.8 trillion parameters, and it comprises 896 individual neural networks, each optimized for a different set of tasks. Harvey trained Tenet on legal documents and added a custom harness, which is a collection of prompts and other technical assets designed to improve output quality. The company says Tenet performs some contract processing tasks 20% better than Kimi K3 and also outperforms Fable 5 and GPT-5 Sol in multiple areas.
Harvey’s platform allows users to store up to 100,000 documents in a repository called Vault, which includes a built-in search engine that uses artificial intelligence to surface useful patterns. For example, an attorney could ask Harvey to find supplier contracts that need modification to comply with a new regulation. Drafting new legal documents often requires lawyers to review existing agreements, and Harvey’s AI helps users locate relevant documents in Vault. It also retrieves external data, such as precedents and legislative clauses, to support that work.
Earlier this year, Harvey rolled out AI agents that enable users to automate more complex tasks, expanding beyond simple document retrieval. An investment firm, for instance, could use these agents to identify potential issues across a large collection of due diligence documents. When the agents need clarification on a tricky part of a task, they ask attorneys for input. This feature is designed to handle workflows that require judgment alongside automation.
Harvey also debuted a benchmark called LAB alongside Tenet, which measures an LLM’s ability to complete legal work. The current iteration of LAB includes about 1,200 tasks, and the company plans to expand it to more jurisdictions and areas of law. Harvey stated in Tenet’s launch blog post that it intends to deploy more computing infrastructure to advance its AI research, with a priority on developing new generalist models. The additional capital should help the company bear the steep costs associated with custom model development.
Training proprietary models could not only help Harvey differentiate its feature set but also improve its margins, as reducing reliance on external models can lower long-term infrastructure costs. This is because inference typically accounts for a much larger share of an AI workload’s cost than training does. The new funding gives the company a larger financial base to pursue that strategy while continuing to serve its legal clients across the U.S.
More company and startup news from TechManNews.
