Datasite

What it is

Datasite is the enterprise virtual data room (VDR) for M&A, run from Minneapolis and controlled by the private-equity firm CapVest. The core products cover the deal lifecycle: Prepare (a free staging room before the deal goes live), Diligence (the sell-side data room), Acquire (buy-side), and Archive. Around them sit Grata for private-market intelligence, Sherpany for board management, and Blueflame AI, the agentic AI company Datasite bought in July 2025. In May 2026 it added Valu8, a financial-data business. Datasite reports 626,000+ users, 16,000+ new transactions a year, and use on 40%+ of the top 100 global M&A deals.

For legal ops the relevant change is that Datasite is now the place your legal AI reads from, not a place you export from. On 2026-04-28 Datasite launched an MCP server that lets Claude, ChatGPT, Microsoft Copilot, and Blueflame create rooms, set permissions, search, summarize change-of-control and termination terms with citations, compare document versions, and draft issue lists, with Datasite’s existing permissions and audit log enforced on every call. On 2026-05-18 Datasite and Legora shipped an integration: Legora users browse Datasite folders inside Legora and import only the files their Datasite permissions allow.

Inside the room, Diligence includes Redaction AI (120+ PII types), full-document translation across 17+ languages, AI summaries, a Q&A workflow with duplicate-question detection, and an index built from a request list. Security credentials: SOC 2 Type II, ISO 27001/27017/27018/27701, and ISO 42001 for AI governance, which Datasite says no other data room holds.

Pricing reality

Datasite publishes no prices. Diligence is quoted per transaction, based on scope, page or data volume, users, and duration. Prepare is free, with no contract or subscription, until you move the project into Diligence.

What buyers pay, per Vendr’s marketplace data checked 2026-10-23: a median of $25,555 a year, range $2,680 to $46,300. Competitor blogs (Papermark, Peony) cite $0.40-0.85 per page and $50,000+ for a mid-market sell-side deal; treat those as directional, since the authors sell against Datasite. The practical read: a contained single-asset sale lands in the low tens of thousands, a broad auction with large Excel models and long timelines climbs from there, and overruns come from page volume and extensions, not seats.

For comparison, Vendr puts the Intralinks median at $13,950 a year (range up to about $150,800). At the low end, Papermark Data Rooms costs $119 a month on annual billing with 3 users, under $1,500 a year, and Peony charges $52 per admin per month. The gap is 10-20x for a team that does not need banker-grade Q&A, support, and certifications.

Best for

Legal ops managers and deal counsel at corporates, PE firms, and law firms who run or support sell-side or buy-side M&A with outside bidders, formal Q&A, and an adviser who expects an institutional VDR, and who want their legal AI (Claude, Legora, Copilot) to work on live room content without downloading it.

Do not use it for fundraising decks, partner document sharing, or a small asset sale with one or two bidders. Papermark or Peony covers that at a fraction of the cost.

Versus the alternatives

• Intralinks (SS&C) — the other half of the institutional VDR market, sold as DealCentre AI. Harvey (2026-06-04) and Legora both integrate with it. Pick Intralinks when your legal AI is Harvey, when your bank or fund already runs on SS&C, or when the quote comes in lower, which Vendr’s medians suggest it often does. Pick Datasite when you want MCP access from Claude or ChatGPT today: Datasite’s April 2026 launch was the first MCP server from a VDR provider.
• iDeals — the mid-market VDR, quote-based across Pro, Business, and Enterprise tiers, with no per-page charges according to iDeals. Pick it when the deal needs a proper Q&A module and audit trail but not Datasite’s buy-side and intelligence suite.
• Papermark — the fastest-growing low-end entrant: open-source core, flat pricing, its own MCP server. Pick it for deals with a handful of counterparties where $1,500 a year beats a five-figure project quote.
• Peony — per-admin pricing for one or two room owners. Pick it when one person runs every room.

Watch-outs

• Page and data volume drive the bill. Guard: get the quote with an explicit cap on pages or GB and on extension fees, and stage in free Prepare first so you know the real volume before signing.
• The MCP server is new and its plan coverage is not public. The launch release names Prepare and Diligence but no tiers or GA status. Guard: write MCP and Legora access into the order form, and run one test query per assistant against a sample room before the room opens to bidders.
• AI access widens who can read the room. An assistant inherits the user’s permissions, so a broad permission group now reaches the AI too. Guard: review permission groups before enabling MCP, and export the audit log of assistant queries at close alongside the Archive copy.
• Blueflame and Grata are upsells. The AI search inside Diligence is not the same as the full Blueflame or Grata products. Guard: list which AI features are in the Diligence quote line by line.

Related on ooligo

Legora, Harvey, Claude for Legal, and Kira Systems cover the legal AI that reads data-room documents.