Spellbook
Last reviewed
Jun 8, 2026
Sources
17 citations
Review status
Source-backed
Revision
v1 · 1,743 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 8, 2026
Sources
17 citations
Review status
Source-backed
Revision
v1 · 1,743 words
Add missing citations, update stale details, or suggest a clearer explanation.
Spellbook is a Canadian legal AI company that builds an artificial intelligence assistant for transactional and contract lawyers. Its core product is a Microsoft Word add-in that lets lawyers draft, review, and red-line contracts, generate and suggest clauses, flag risks, benchmark deal terms against market data, and answer questions about a document, all without leaving the word processor where most contract work is done [1][2]. The company was an early adopter of large language models in law, launching what it has described as one of the first generative AI tools built specifically for lawyers in late 2022, powered at the time by OpenAI's GPT-3 [3][4].
Originally created under the company name Rally (also known as Rally Legal), Spellbook is headquartered in St. John's, Newfoundland and Labrador, and is led by co-founder and chief executive Scott Stevenson [5]. By late 2025 it reported nearly 4,000 law firms and in-house legal teams as customers across roughly 80 countries, including Nestle, eBay, and the international law firm Kennedys [2][6]. The company has raised more than $80 million USD in equity across seed, Series A, and Series B rounds, and in March 2026 added a $40 million USD debt facility from RBCx earmarked for acquisitions [2][6][7].
The company was founded in 2018 in St. John's, Newfoundland, as Rally, a document automation and templating engine for law firms and small businesses, by Scott Stevenson together with co-founders Matt Mayers and Daniel Di Maria [5][8]. Rally raised roughly $750,000 in pre-seed funding in 2019 [8].
Stevenson has said the idea for Spellbook grew out of his own experience: legal fees consumed a large share of an earlier startup's investment, which motivated him to build software to reduce the cost and drudgery of contract work [9]. Spellbook itself began as a marketing experiment inside Rally, conceived as a way to generate leads for the document automation business [10]. After OpenAI made its language models available, the team fine-tuned the technology on legal data and built a Word add-in that could suggest contract language in context. Rally publicly unveiled Spellbook in November 2022, describing it as the first AI contract drafting tool powered by GPT-3 [3][4].
The Spellbook product proved far more popular than the underlying Rally automation tool, and the company formally rebranded from Rally to Spellbook, adopting the name of its breakout product [1][5]. Stevenson became the public face of the company as CEO, with Daniel Di Maria serving as chief operating officer [5]. Spellbook secured early access to successive OpenAI models, including GPT-4 and later GPT-5, and in August 2025 was among the first legal AI platforms to ship a product built on GPT-5 [4][11].
Spellbook's flagship product is a sidebar add-in that runs inside Microsoft Word, the application where the bulk of transactional drafting and negotiation already happens. This Word-native approach is central to the company's positioning: rather than asking lawyers to move their work into a separate web platform, Spellbook embeds AI directly into the contract document. The original add-in remains what most users rely on [1].
Within the add-in, Spellbook can review a contract for risks and drafting errors, suggest and generate clauses, draft full sections or documents from scratch, summarize provisions in plain language, and propose negotiation points. The product offers several review modes, including a general review that scans broadly for issues, a negotiate review tuned to favor the party the lawyer represents, and narrower custom reviews [1]. A Playbooks feature lets firms encode their own negotiation standards and positions into reusable rule sets so the AI applies consistent guidance across deals [1].
In August 2024, Spellbook launched Spellbook Associate, which it positioned as a more autonomous, agentic product for longer multi-document transactional workflows. Rather than acting as a single-prompt chatbot, Associate plans a sequence of steps, retrieves relevant precedents, drafts and revises clauses, updates party details and defined terms across a set of documents, and can assemble a complete package such as a financing deal from a single term sheet, validating intermediate outputs along the way [12][13]. It is aimed at transactional areas such as mergers and acquisitions, real estate, and business formation [13]. The company has compared this agentic direction to coding assistants like Cursor, framing the goal as a "Cursor for contracts" [10].
More recently, Spellbook has pushed into grounding its outputs in real contract data. In early 2026 it introduced a Compare to Market capability that benchmarks a deal's terms against a large corpus of comparable agreements so lawyers can see how their positions stack up against prevailing market practice [2][14]. The company has said it matches contracts against thousands of contract types and draws on data from hundreds of thousands of agreements to inform these comparisons [2][14].
Spellbook has raised both equity and debt as the legal AI market expanded. The figures below are drawn from company announcements and reporting, and the round labels follow how the company and press have described them.
| Round | Date | Amount | Lead investor | Notes |
|---|---|---|---|---|
| Pre-seed | 2019 | ~$0.75M USD | (as Rally) | Document automation startup [8] |
| Seed | May 2023 | $10.9M USD | Moxxie Ventures | Rebrand to Spellbook; Thomson Reuters Ventures and others participated [1][8] |
| Series A | January 2024 | $20M USD (~$27M CAD) | Inovia Capital | Total funding then exceeded $30M; ~1,700 customers [8][15] |
| Series B | October 2025 | $50M USD (~$70M CAD) | Khosla Ventures | $350M post-money valuation; total raised over $80M [6][16] |
| Debt facility | March 2026 | $40M USD | RBCx (debt) | Earmarked for acquisitions [2][7] |
The May 2023 seed round of $10.9 million USD was led by Moxxie Ventures, with participation from investors including Thomson Reuters Ventures, Inovia Capital, The LegalTech Fund, and Bling Capital, and coincided with the rebrand from Rally to Spellbook [1][8]. In January 2024, Spellbook raised a $20 million USD Series A (about $27 million CAD) led by Inovia Capital, bringing total funding past $30 million; at that point the company reported more than 1,700 law firms and legal teams as customers and more than 86,000 contracts opened on the platform each month [8][15].
On October 9, 2025, Spellbook announced a $50 million USD Series B (roughly $70 million CAD) led by Khosla Ventures, with managing director Keith Rabois joining the board. The all-equity round valued the company at $350 million post-money and included participation from Threshold Ventures and returning investors such as Inovia Capital, Bling Capital, Moxxie Ventures, Path Ventures, and former Shopify chief technology officer Jean-Michel Lemieux. The round brought total equity raised to more than $80 million [6][16]. At the time, Spellbook said it planned to grow from about 115 employees, roughly 110 of them in Canada, to around 230 over the following one to two years, and to expand beyond contract review into broader transactional work [16].
On March 4, 2026, Spellbook secured a $40 million USD debt facility from RBCx, the technology and innovation banking arm of the Royal Bank of Canada (RBC), specifically to fund strategic acquisitions in what it described as a rapidly consolidating legal AI market [2][7]. Stevenson said the company was being approached by a small legal AI company looking for an exit every few weeks, and that the market needed consolidation [2][17]. The company said 2025 had been a landmark year in which it tripled revenue [2][7].
Spellbook targets small and midsize law firms and in-house legal teams, a segment it argues is underserved by enterprise-focused platforms, although by late 2025 it reported that a substantial share of its revenue came from larger firms and enterprise in-house teams as well [1][6]. Its differentiator is the Word-native delivery model: meeting lawyers inside the tool they already use rather than requiring adoption of a separate platform.
The legal AI market it competes in has attracted heavy investment and several billion-dollar valuations. Broad legal AI platforms such as Harvey and Legora raised large rounds at high valuations over 2025 and 2026, and incumbents including LexisNexis and Thomson Reuters (through its CoCounsel assistant) compete with full research-and-drafting suites [2][17]. Among tools focused, like Spellbook, on contract drafting and review, competitors include Robin AI, LegalOn, and Gavel Exec, while contract lifecycle management vendors such as Ironclad, Icertis, and Docusign address adjacent parts of the workflow [1].
Stevenson has framed Spellbook's strategy around two ideas: grounding AI output in real contract data to make it more useful and trustworthy for negotiation, and using its capital, including the RBCx debt facility, to acquire complementary technology and talent as smaller players seek exits. As of mid-2026 the company positioned itself as one of the larger independent legal AI vendors, with nearly 4,000 customers across about 80 countries and more than 10 million contracts reviewed on its platform [2][6][16].