Sales Leader

Harvey Tenet: why enterprise AI apps must own their models. On August 18, 2026, legal AI company Harvey crossed a line that most enterprise AI software vendors have not reached: it released its first proprietary large language model. Harvey Tenet, post-trained on the open-source Kimi K3 foundation using years of legal reasoning data, marks the transition from Harvey as an API integrator to Harvey as a model owner. The shift carries consequences well beyond one legal tech company. The model launch is part of a broader platform release Harvey calls Harvey II, which also includes persistent Memory across matters, a redesigned Spaces interface, and updated agent workflows. Together, these changes are the clearest signal yet that vertical enterprise AI companies are entering a new strategic phase. The wrapper problem at scale. Harvey built an $11 billion legal software business by combining models from OpenAI, Anthropic, and others into a platform that helps lawyers accelerate document review, drafting, and research. Each query routes through a third-party API call billed per token. At small usage volumes, that economics holds. At the scale Harvey now operates, the math turns. Every new lawyer seat, every additional matter, every background agent task adds to the external model bill. A software company paying a per-token tax on its own revenue has a structural ceiling on gross margin that a model owner does not. Tenet is Harvey's answer. According to the official announcement, Tenet is "frontier-level on prominent legal benchmarks, performing on par with the strongest general models at an open-source cost." That pairing, frontier accuracy on domain benchmarks at open-source inference cost, is the business thesis in a single sentence. Harvey is not claiming Tenet outperforms GPT-5.6 Sol on general reasoning. It is claiming Tenet matches the best general models on legal reasoning, which is the only dimension that matters for Harvey's product, at a fraction of the token price. How Tenet was built. Harvey trained Tenet on Kimi K3, the open-weight frontier model released by Moonshot AI in July 2026. Kimi K3 attracted enterprise interest precisely because it offered competitive performance under a permissive license, making it viable as a base for proprietary fine-tuning without the IP friction of building on closed models. Harvey added its proprietary dataset, built by lawyers who generate and evaluate outputs, to post-train Tenet for legal reasoning end-to-end. That dataset is the actual moat. Any company can download Kimi K3. Only Harvey has years of structured legal reasoning data tagged by practicing attorneys. This is the architecture that makes vertical AI defensible at scale: an open-weight foundation plus proprietary domain data. The foundation manages training cost. The data creates the differentiation that cannot be cheaply replicated. It also removes a dependency risk. Harvey no longer needs OpenAI or Anthropic to keep Tenet running on core legal workflows. Those frontier API relationships can continue for tasks where general capability matters, but the critical path through Harvey's product no longer routes exclusively through a supplier that can change pricing, access terms, or model behavior unilaterally. Harvey II: memory, Spaces, and continuous context. Beyond Tenet, Harvey II solves a problem that has frustrated enterprise AI deployments since the first wave of agentic tools: agents that begin every task without memory of previous work. Memory carries a lawyer's drafting style, citation preferences, matter history, and approved best practices across tasks. Phase one serves individual users. Phase two, rolling out over the following months, extends memory across Spaces and team matters. Phase three allows firm-wide governance of memory boundaries, including what gets stored and what remains scoped to individual matters. Spaces creates a matter-centric shared environment where documents, tasks, permissions, and history stay attached to the work as it moves between agents and lawyers. Rather than a single prompt-response session, Spaces is designed to support continuous multi-session workflows across an entire matter lifecycle. According to Harvey's CPO Anique Drumright, Memory came directly from lawyers describing how much time they wasted re-explaining preferences to an AI that had forgotten everything since the last session. "Instead of repeatedly explaining those preferences or spending time reshaping outputs afterward, Harvey can start closer to the way that lawyer actually works." | Harvey II Feature | What It Does | Enterprise Relevance | | Memory (Phase 1) | Stores individual style, tone, citation preferences | Reduces per-task setup overhead | | Memory (Phase 2) | Extends across Spaces and shared matters | Enables continuity on long-running projects | | Memory (Phase 3) | Firm-wide memory governance and boundary controls | Addresses compliance and confidentiality | | Spaces | Matter-centric environment with persistent history | Replaces scattered document and prompt workflows | | Harvey Tenet | Proprietary legal model at open-source inference cost | Removes per-token external cost on core workflows | What this means for enterprise AI strategy. Harvey's move illustrates a transition that many enterprise AI companies will face in the next 12 to 24 months. The dynamic follows a recognizable pattern. Phase 1 (Wrapper). The company integrates frontier model APIs to ship a vertical product quickly. Margins are acceptable because usage is modest and the product value justifies the token cost. Phase 2 (Scale pressure). As usage grows, the per-token cost becomes a meaningful line in the P&L. The company finds itself paying a supplier tax on every dollar of revenue. Phase 3 (Model ownership). The company fine-tunes an open-weight model on its proprietary data. External API calls shift to tasks where general capability genuinely matters; core workflows route through the internal engine. Harvey is entering Phase 3. The timing is notable: at an $11 billion valuation, not at inception. The capital and data required to reach Phase 3 are only available to a company that has already proven market fit and accumulated substantial domain data. Earlier-stage vertical AI companies should treat the Harvey Tenet launch as a milestone that defines what they are building toward. Harvey's longer-term vision goes further. CEO Winston Weinberg has described Tenet as a potential "building block" that law firms use to train their own models shaped by their own legal work. If that materializes, Harvey moves from a software supplier to a model foundry for the legal industry: individual firm models, each differentiated by proprietary matter data, with Harvey providing the common base and infrastructure layer underneath. That is a very different business than an API wrapper, and it is the direction the enterprise AI software market is heading. The companies that reach that destination will have done so because they accumulated irreplaceable domain data while others were focused solely on the interface layer. For enterprise teams assessing vertical AI vendors, the question is no longer whether a vendor uses frontier models. The question is whether the vendor is building its own data flywheel and model ownership path. Vendors with no answer to that question are structurally exposed to margin compression and supplier dependency as their usage scales. Understanding model strategy at the platform level is one of the decisions that separates AI-native enterprises from those still experimenting. For teams navigating that transition, Enera works with enterprise buyers and operators to build durable AI capabilities that do not rent their core intelligence from a changing market.

Harvey taps DeepL to power legal-grade AI translation for global law firms and legal teams. DeepL supports Harvey's enterprise translation engine, powering highly secure, context-aware document translation and handling over a third of Harvey's total volume SAN FRANCISCO, Aug. 19, 2026 /PRNewswire/ - DeepL, a global leader in Language AI, today announces that Harvey, the leading AI platform for law firms and legal teams, has selected DeepL to power fast, accurate and secure document translation directly inside its platform. The integration reflects growing demand for specialized AI tools built to handle the complexity of legal work, where accuracy, speed, formatting, terminology and security matter. "Harvey customers work with complex, high-stakes documents all the time and document translation is one of our most-used workflows" said Lauren Oh, Product Manager at Harvey. "By integrating DeepL's AI document translation capabilities directly into the Harvey platform, we're giving legal teams an even faster, more seamless and reliable way to work across languages, while also making sure content stays highly precise and accurate." Harvey serves more than 200,000 lawyers across 2,400+ organizations in 70 countries, including many of the world's largest law firms and in-house legal teams. For many of these customers, language is not just a side issue but a daily challenge. Cross-border legal work depends on translating contracts, filings, briefs, evidence, reports, and client materials quickly, accurately and reliably even when documents are long or highly complex. With DeepL embedded directly into the Harvey platform through its translation API, customers can now upload any document, select a target language, and receive a precise, customized translation that preserves its original structure and formatting, all without ever leaving Harvey. The integration currently supports over 100 languages, and will handle over a third of Harvey's total document translation volume. "Working with Harvey brings our specialized Language AI capabilities to even more legal teams around the world, who work across borders every day," said Jarek Kutylowski, Founder and CEO of DeepL. "Legal work usually means dealing with a lot of documentation, whether that's contracts, filings or case documents that span hundreds of pages, in dozens of different formats and unique contexts. Our platform is built for exactly this level of complexity, and Harvey choosing us for it is clear validation of DeepL's leadership in the Language AI market." DeepL was selected as one of Harvey's AI translation vendors for its ability to preserve the structure and formatting of complex legal documents, while also delivering context-aware translations powered by specialized language models purpose-built for translation. DeepL's custom glossaries functionality and unmatched file-format support help legal teams maintain approved terminology and document integrity across languages, no matter the topic, complexity or length. The company's strong security and compliance standards, including no permanent data retention for enterprise and paid users, GDPR compliance, SOC 2 Type II, ISO 27001 certification, and more, are also central to this work and Harvey's requirements for handling confidential and high-stakes materials. This news builds on DeepL's broader momentum worldwide as businesses increasingly embrace Language AI as a core infrastructure for global work. DeepL's platform spans written and spoken translation, from text and documents to real-time translation for virtual meetings, in-person conversations and live events, and is used today by over 200,000 business teams and millions of individuals across nearly every industry, from legal to healthcare, manufacturing, financial services and technology. It also comes as DeepL continues to expand in the US, where nearly 50% of the Fortune 500 are users. About DeepL DeepL is a global AI company building the language infrastructure that powers global business. More than 200,000 business teams and millions of individuals use DeepL's Language AI platform to communicate globally, collaborate and operate across languages in real time. By combining breakthrough AI models with enterprise-grade security and privacy, DeepL enables organizations to work seamlessly across markets and cultures. Founded in 2017 by CEO Jarek Kutylowski, DeepL now has more than 900 employees and is backed by leading investors including Benchmark, IVP and Index Ventures. Learn more at www.deepl.com. About Harvey Harvey is the operating system for legal and professional services. Our products streamline workflows in areas including contract analysis, due diligence, compliance, and litigation to drive efficiency and value. Global law firms and Fortune 500 enterprises around the world use Harvey to enable faster, smarter decision-making. Backed by world-class investors including Sequoia, Kleiner Perkins, GV, OpenAI Startup Fund, Coatue, Andreessen Horowitz, GIC and EQT, Harvey is used by 2,400+ customers in 70+ countries. SOURCE DeepL

Legaltech giant Harvey launches next-generation AI platform. Harvey II brings together features including personalisation tool Memory 18 August 2026 Harvey's chief product officer Anique Drumright Legaltech giant Harvey has announced the launch of Harvey II, creating a platform that promises to accelerate personalised AI work for lawyers and move towards a greater focus on agentic legal tasks. The firm says the new platform brings together features including Memory, an AI lawyer personalisation tool that enables individual preferences to be carried across Harvey, Word, Outlook and agents. The platform incorporates a 'spaces' technique where documents, tasks, permissions and matter or project history can be brought together in one place. Smarter agents can then work with that context already in place, alongside a refreshed product experience assembled around how the work progresses. Anique Drumright, chief product officer at Harvey, said: "Memory is a major differentiator for Harvey and a key part of making AI more personalised to the individual lawyer. "It allows Harvey to learn and retain how a lawyer works, from drafting style and preferred structure to citations, level of detail and other preferences developed over the course of working with Harvey." Advertisement Drumright says Memory is a vital facet of the tool, as legal professionals can approach the same assignment differently. Lawyers can avoid having to repeatedly explain their preferences, as Memory knows how they like to work in context, in a move away from AI responding to individual prompting. Harvey says Memory is built around user control and transparency, enabling lawyers to see what Harvey remembers, while also being able to change what it recollects. Drumright explained: "For example, a lawyer could return to an M&A matter and have Harvey already worked through the latest material contracts, providing a first-pass review in their preferred format with findings cited, while surfacing the agreements that need attention, and then open the next step for the lawyer responsible." The goal is to ensure less time is spent explaining preferences, finding documents and rebuilding context, and more time on substantive legal work, the firm said, adding that Memory was designed in response to customers wanting a more bespoke legal AI offering. "For us, that's an important part of leading this market: building with our customers and delivering the capabilities that have the greatest impact on their work," Drumright said. The launch of Harvey II intensifies its agentic approach to its legal AI products, performing multi-step tasks across case processes. LAW OVER BORDERS COMPARATIVE GUIDES Artificial Intelligence Law Guide This second edition, written by leading AI legal specialists, provides answers and insight on how to integrate Artificial Intelligence into business operations, whilst working within the relevant law and guidelines in key jurisdictions around the world... | 1yr. Over the past six months since Memory was first announced, Harvey has hosted 'listening tours' spanning the US, UK, Europe, the Middle East and Asia-Pacific. A cohort of global law firms and in-house legal departments, including AmLaw and Magic Circle firms along with Fortune 500 and FTSE 100 legal departments, have been considered design partners for Memory. Drumright commented: "Our listening tours and work with law firms and in-house legal teams as design partners helped us understand what Harvey should remember, how those memories should be applied, and where they need to stop. "That input has been critical to building personalisation that becomes more useful over time while respecting the permissions, confidentiality requirements and ethical walls that legal work demands." Memory is being rolled out in three phases, the initial step - 'personal' - is currently in an early access stage, focused on lawyers retaining preferences around such issues as drafting. In the next few months, the 'space' phase will be introduced, where relevant 'memories' will be applied across a shared area. The final phase will be at an organisation level, where firms and legal departments can bring broader ways of working into Harvey, sustaining appropriate boundaries between users. Harvey says that Memory will reach a general access stage for clients during the third quarter of this year.