Processes 1M+ daily legal queries
Analyzes 115M+ legal documents daily
Empowers 200,000+ users across 85+ countries
Executes long-horizon legal agent workflows
Scales regional cloud capacity in minutes
Harvey AI is transforming legal workflows for over 200,000 professionals worldwide by using Claude models on Google Cloud to deliver secure, high-scale legal analysis.
Harvey AI is transforming legal workflows for over 200,000 professionals worldwide by using Claude models on Google Cloud to deliver secure, high-scale legal analysis.
Reimagining legal workflows under strict security and scale demands
Founders Winston Weinberg and Gabe Pereyra built Harvey AI on a unique goal: blend deep legal domain expertise with groundbreaking AI. Today, Harvey serves over 3,500 customers across more than 85 countries and has become a valuable daily platform for over 200,000 legal professionals. Whether analyzing complex contracts, drafting briefs, or managing due diligence, lawyers use Harvey to accelerate critical workflows without compromising security or accuracy.
Building an AI platform for the legal industry presents technical and operational hurdles. Legal processes rely on handling massive, complex documents. In addition, law firms and corporate legal departments have to manage confidential information without the chance for any security breaches.
Harvey needed to enforce strict zero data retention, maintain localized data residency, and guarantee high platform availability during usage spikes. To satisfy attorneys who need advanced performance and security, Harvey sought a cloud and AI architecture that could scale across global regions while adhering to privacy controls.
Lawyers deal with confidential documents and care about security, zero data retention, and data residency. To earn their trust, we needed partners like Google Cloud and Anthropic who shared our strict security mindset as we scale rapidly.
Gary Lam
AI Platform Lead, Harvey
"Lawyers deal with confidential documents and care about security, zero data retention, and data residency,” says Gary Lam, AI Platform Lead at Harvey. “To earn their trust, we needed partners like Google Cloud and Anthropic who shared our strict security mindset as we scale rapidly."
Orchestrating frontier intelligence across Google Cloud and Anthropic
Harvey worked with Google Cloud and Anthropic to power its specialized platform, leveraging Model Garden to access industry-leading models, from both Anthropic and Google.
Every legal task is unique. Having Claude on Google Cloud lets us optimize and serve the best model for a lawyer's specific workflow, from contract redlining to complex analysis.
Gary Lam
AI Platform Lead, Harvey
This infrastructure allows Harvey to deploy localized application instances directly within geographic regions like the EU, US, and Australia, guaranteeing that sensitive customer data is processed locally to meet data residency laws.
Behind the scenes, Harvey uses a sophisticated auto-orchestration layer and model selector that pairs legal tasks with the ideal model architecture. Legal tasks range from a single redline edit to multi-step reasoning across thousands of documents.
Attorneys can interact with these tools directly within their everyday workflow through Microsoft Word, Outlook, and mobile applications. Internally, Harvey even deployed Spectre—a custom cloud-based internal coding agent that utilizes Model Garden—so engineers, product managers, and designers can collaborate and accelerate product development.
"Every legal task is unique,” says Lam. “Having Claude on Google Cloud lets us optimize and serve the best model for a lawyer's specific workflow, from contract redlining to complex analysis."
Unlocking long-horizon reasoning and daily scale for global law
By anchoring its architecture on Google Cloud, Harvey has achieved exceptional scale, processing over 1 million queries and more than 115 million documents every day. Accessing Claude models using Google Cloud has dramatically expanded the platform's reasoning capabilities, enabling autonomous AI agents to handle long-horizon tasks across large documents spanning hundreds of pages without jeopardizing accuracy.
"Harvey shows what’s possible when frontier model choice meets enterprise-grade cloud infrastructure,” says Shikhar Mathur, Global Partner Manager, Google Cloud. “By combining Anthropic’s Claude and Google’s Gemini models on Google Cloud, Harvey delivers high-precision legal agents across millions of documents while upholding the strict security, zero data retention, and regional compliance law firms require."
Google Cloud's reliable infrastructure also empowers Harvey to instantly provision dedicated capacity in international regions, maintaining high responsiveness during demand surges.
Working with Google Cloud and Anthropic lets us reliably scale worldwide in minutes. We can give our long-running legal agents a longer leash, delivering higher-quality outputs on complex, multi-page tasks.
Gary Lam
AI Platform Lead, Harvey
Looking ahead, Harvey continues to redefine legal tech with Harvey II, which gives agents memory of lawyers’ preferences and the context of their matter or project. Supported by Google Cloud and Anthropic, Harvey remains at the forefront of AI-driven legal transformation.
"Working with Google Cloud and Anthropic lets us reliably scale worldwide in minutes,” says Lam. “We can give our long-running legal agents a longer leash, delivering higher-quality outputs on complex, multi-page tasks."
Founded by legal and AI pioneers, Harvey provides an advanced platform designed to accelerate legal research, document review, and contract drafting for global law firms and corporate legal departments. Today, the platform serves over 200,000 legal professionals across more than 85 countries, enabling seamless collaboration and automated workflows for complex legal matters.
Industry: Technology
Location: United States
Products: Gemini, Claude on Google Cloud
About Google Cloud partner - Anthropic
Anthropic is an AI safety and research company dedicated to building reliable, interpretable, and steerable frontier AI systems that power complex enterprise tasks.
