Spellbook

What it is

Spellbook is contract AI for transactional lawyers. It started as a Microsoft Word add-in that suggested redlines in the sidebar, and that add-in is still the daily surface, but the product sold today is a platform: it reviews incoming paper against your standards, drafts from your own precedent, compares terms to market, answers questions across a document set, and encodes house positions as playbooks. It reports 4,500+ legal teams across 80+ countries, including Dropbox’s in-house legal team and the firm Kennedys. An October 2025 Series B led by Khosla Ventures put $50M in at a $350M post-money valuation, taking total funding past $80M, followed by a $40M RBCx debt facility in March 2026 earmarked for acquisitions.

Two things changed in 2026 that you need before shortlisting it. The company moved from spellbook.legal to spellbook.com — the old domain now 301-redirects — and it stopped publishing prices entirely.

Pricing was withdrawn, not lowered

Spellbook used to publish Solo at $99/user/month and Team at $179/user/month. Those tiers are gone from the site. spellbook.com/pricing now shows two segments, Law Firms and In-House Teams, with no dollar figure anywhere on the page, and states that pricing is determined by the number of team members on a license. The only self-serve door left is a 7-day free trial; everything else routes to a demo. Dedicated support is marked as a benefit for teams over 10.

Treat any $99 you find in a review article or an old procurement sheet as a historical vendor price, not a current quote. Third-party trackers have filled the vacuum and disagree wildly: published estimates put the entry seat anywhere from roughly $20 to $99 per user per month and enterprise from roughly $199 to $350, with a reported 10-seat and six-month minimum. A 5x spread on the entry tier is not a price band — it is evidence that nobody outside Spellbook’s sales team knows. The one directional signal worth carrying into a budget is that the reported enterprise seat sits near double the $179 Spellbook last published for Team.

Why it shows up in Legal Ops stacks

• It meets lawyers in Word. No migration, no new system of record. A ten-lawyer transactional team can be redlining on day one, against a 3-6 month implementation for Ironclad or a full CLM.
• It reads your paper, not just the market’s. The iManage connector pulls answers out of your own document history, and OneDrive, SharePoint, Google Drive, and Dropbox connectors do the same for teams that keep contracts in cloud storage. A separate legal-data connector reaches 200+ public legal databases to ground citations.
• Playbooks make the review repeatable. Encoding fallback positions once means first-pass review runs the same way whether the deal lands with the GC or a contract manager — the specific case where in-house teams get the most out of it is high-volume vendor paper, NDAs, and MSAs where the standards are settled and the volume is the problem.
• Associate widens the unit of work. Where Review works one document, Associate handles multi-document jobs — working a term sheet up into a full financing set, or running inconsistency checks across hundreds of files.

Best for

• In-house counsel and Legal Ops managers handling vendor contracts, NDAs, and MSAs at volume, who want first-pass review off the lawyers’ desks
• Small and mid-market transactional practices — corporate, commercial, employment — whose work already lives in Word
• Teams whose first AI rollout has to show a result inside a quarter rather than a fiscal year

Watch-outs

• You are entering a sales conversation, not a checkout. Budget the procurement time you would have skipped under the old $99 self-serve tier, and get the quote in writing before modelling seats. Guard: run the 7-day trial first so the pilot result is in hand before the pricing call, and re-check the number at renewal — this vendor moved from published to quote-only inside a year.
• Word-first is still a real constraint. If the firm has standardized on Google Docs for drafting, the sidebar is not where your lawyers work. Guard: confirm the drafting surface, not just the storage connector — a Google Drive integration reads your files, it does not put Spellbook inside Google Docs.
• It does not do primary-law research. Spellbook draws authority from the contract in front of you, not from a case-law corpus, and does not claim otherwise. Guard: anything headed for a filing still needs a KeyCite or Shepard’s check in Thomson Reuters CoCounsel or Lexis+ with Protégé.
• Autonomous Contract Management is announced, not delivered. ACM — autonomous intake from email, Slack, and Salesforce, plus post-signature storage and renewal flagging — was announced on June 30, 2026 and is rolling out to selected teams from a waitlist. Guard: buy Spellbook for the review and drafting that ships today; if the CLM replacement is the reason you are buying, get the rollout date and the metering in the contract, because agentic monitoring across tens of thousands of contracts is exactly where the token cost is unproven.

Alternatives and when to pick them instead

Harvey is the enterprise incumbent and wins on governance depth and breadth of workflow — pick it if you are an AmLaw firm or a large enterprise legal department where research, matter work, and contracts all need one vendor. Ironclad wins when the actual problem is the contract lifecycle rather than the drafting: approvals, repository, and renewals as a system of record. Legora is the fastest-growing entrant in this segment and is the one to test against Spellbook if a collaborative agent workspace, rather than a Word sidebar, is where you want the work to happen.

If the quote comes back above budget, Ivo, Definely, and DraftWise attack review and drafting from different angles at a smaller footprint — Spellbook vs DraftWise works that trade in detail.