NetDocuments wants its document system to be the context layer under Harvey, Legora and ChatGPT. Here is what is live, what is in preview and what it costs.
NetDocuments has sold cloud document management to law firms for more than 25 years. This fall it is selling a bigger idea: that the document management system should become the firm’s “intelligence layer,” the governed store of context that every AI tool draws on, whichever tool a lawyer happens to like this month. Its Inspire conference in Orlando last month and a new LawNext podcast with its CEO make this a good time to look at what is shipping and what buyers can check.
What It Claims and Ships
NetDocuments says it supports more than 7,000 law firms, corporate legal departments and public sector organizations, and integrates with more than 150 other technologies, including Microsoft 365, Claude by Anthropic and Docusign. Those are the company’s own figures, from its October 15 release.
The AI products sit under the ndMAX name. That October release added six apps to ndMAX Studio, bringing the catalog to 45 pre-built apps, with new ones for family law, estate planning, fund formation, venture financing and corporate governance. Examples include a Divorce Financial Disclosure Analyzer, a Will and Trust Internal Consistency Checker and a Corporate Actions Register App that turns board minutes and resolutions into a searchable register. Firms can also build their own apps with the ndMAX AI App Builder.
In March, the company launched Smart Answers, natural language questions answered from a firm’s own repository with citations, rolling out to ndMAX Enterprise customers from March 31. The same release announced MCP connectivity so that ChatGPT, Claude and other compatible tools can reach NetDocuments content under existing permissions and ethical walls, available to ndMAX Enterprise customers from April 1. Chief Product Officer Dan Hauck framed the pitch this way:
“Any legal professional can open a public LLM and ask a question. But the answer they receive is the same answer every other firm gets, because those models draw from the same generalized training data,” Hauck said.
The bigger bet is what NetDocuments calls the first Legal Context Graph. LawSites reported in May that the graph maps relationships among matters, documents, communications and people at the document, matter and firm level, built with AWS and Elastic and drawing in part on SALI Alliance and FOLIO taxonomy standards. It arrived with a redesigned interface that includes a generated matter overview, natural language search, AI summaries of what changed between versions and an AI editor inside Word. Customers can toggle between the new and old interfaces on the same data.
Status matters here. In May, the graph and new interface opened as a private preview for select enterprise AI tier customers. In its own Inspire recap on October 23, NetDocuments still described the graph as in private preview. The same recap says Legora showed a new MCP connection to NetDocuments on stage, and that speakers from Harvey and OpenAI joined a session on using NetDocuments context over MCP.
On the LawNext podcast published October 30, CEO Josh Baxter told Bob Ambrogi that NetDocuments does not build models, that users will swap the “harness” at will, and that the firm’s own data and context is the layer that will last. LawSites’ summary says he even acknowledged that the DMS may eventually become “headless.”
What the Public Evidence Shows
Most of the numbers come from NetDocuments. Its recap says an internal Legal Context Engineering Benchmark, which held the model, tool and questions constant, found that building context once cut the cost of each correct answer in half on a top-tier model with no loss in quality. Baxter told LawSites the savings reach about 70% on the most expensive models, and he reported month-over-month AI usage growth of 40% to 50% on the platform. The March release said more than 800 firms began using NetDocuments AI in 2025. None of that has been independently tested in anything we found.
What is better documented is the direction of the market. Harvey, Legora and OpenAI are connecting to the document system rather than asking firms to upload everything again, and the recap says each MCP request runs under the lawyer’s own NetDocuments access, with permissions, ethical walls and audit trails applied. That is the right design for a governance team. It also means the value depends on how well a firm’s documents are already profiled and organized.
Who It Fits
NetDocuments fits firms and legal departments that want one governed store of documents and email and expect lawyers to use several AI tools on top of it. Firms already on NetDocuments have the easiest path, since the company says Smart Answers can be switched on without a new system or a separate contract, though the AI features named above are tied to the ndMAX Enterprise tier.
Pricing is not posted. The pricing page says cost depends on team size, products, security needs and migration, and asks buyers to fill out a form. Ask for an itemized quote that separates the DMS from ndMAX, ask whether AI can be licensed to a subset of users, and ask when the context graph leaves preview.
Common Questions
Is the Legal Context Graph generally available?
Not in what we found. NetDocuments opened a private preview in May and still called it private preview in its October 23 recap.
Can firms use Harvey, Legora or ChatGPT with NetDocuments?
NetDocuments said MCP connectivity to ChatGPT and Claude would reach ndMAX Enterprise customers from April 1. At Inspire, Legora demoed its MCP connection and Harvey discussed using NetDocuments context over MCP.
Does content leave NetDocuments when AI tools connect?
NetDocuments says MCP requests run under each user’s permissions and ethical walls with an audit trail, and its March release said no content leaves the platform.
How much does NetDocuments cost?
The company does not publish prices. Quotes depend on users, products, security requirements and migration.
The Verdict
Partly Real. Smart Answers, ndMAX Studio and MCP connections are shipping features, and the case for governing context once is sound. The headline graph is still in preview and the savings figures are the company’s own, so ask for a live demo on your own matters before you plan around them.
