Spellbook
Contract drafting and review inside Microsoft Word, with the agentic Spellbook Associate review layer. Self-serve per-seat for 4,000+ legal teams.
Spellbook handles contract drafting and review inside Microsoft Word, with 'Spellbook Associate' as its agentic review layer, serving 4,000+ legal teams and backed by a $50M Series B. Its self-serve per-seat model fills the mid-market gap that Harvey's seat minimums exclude; pricing runs from $99/user/month, with Pro tiers around $329–399/user/month.
Vendor details
Canonical URL
https://spellbook.legal
Category
Enterprise operations agent
Subcategory
Legal — contract drafting/review (Word add-in)
Use cases & customers
Primary use cases
Target customers
Deployment options
In practice
Your lawyers live in Microsoft Word, not another legal app. Spellbook works inside Word, handling contract drafting and review where the document already is.
A contract needs a full review and the associate who'd do it is slammed. Spellbook Associate runs an agentic review pass, surfacing issues before a human takes the final look.
Enterprise legal AI won't sell to a small team without a big seat minimum. Spellbook is self-serve and per-seat, built for the mid-market firms those minimums exclude.
Agentic Index coverage score
5.5 / 14 capabilities · 39%
| Integrations & Tool CallingSpellbook runs natively inside Microsoft Word and Associate works across document sets and data room materials, but the product is designed exclusively for Word on desktop with no support for other document editors and no documented third party tool calling, Spellbook review 2026-06-04 and in-house counsel review 2026-07-22 | Partial |
|---|---|
| Workflow OrchestrationSpellbook Associate plans and executes multi document projects, taking a goal and breaking it into steps completed in a multi stage workflow across different documents, and can plan, execute, check its work, and adapt to accomplish larger scope assignments, Spellbook Associate product page 2026-07-22 and Associate launch coverage 2024-08-22 | Full |
| Knowledge Grounding & RAGDrafting is grounded in the customer's own precedents and reusable playbooks, and Compare to Market benchmarks terms against an industry dataset segmented by sector, jurisdiction, and deal type, with an access model where teams contribute anonymized contract data to unlock the benchmarking pool, Spellbook in-house counsel review 2026-07-22 | Full |
| Human Oversight & GuardrailsPlaybooks codify a team's negotiation standards so agent behavior is bounded by firm policy and output lands as redlines a lawyer accepts in Word, but Spellbook explicitly frames Associate as working without constant supervision and no approval gates were documented, Spellbook Associate launch coverage 2024-08-22 and product pages 2026-07-22 | Partial |
| Security, Identity & GovernanceNo security attestations, identity controls, or agent permission model was retrieved in this pass, Spellbook product pages 2026-07-22 | Unable to verify |
| Observability & AuditabilityNo agent run tracing, execution logging, or decision audit trail was retrieved in this pass, Spellbook product pages 2026-07-22 | Unable to verify |
| Memory & State PersistenceNo agent memory or cross session state mechanism was retrieved in this pass, Spellbook product pages 2026-07-22 | Unable to verify |
| Deployment & Data ResidencyNo deployment options, regional hosting, or data residency controls were retrieved in this pass, Spellbook product pages 2026-07-22 | Unable to verify |
| Prebuilt Agents, Templates & PacksThe product ships a named capability set out of the box covering Review for redlining and risk flagging, Draft, Ask, Compare to Market, and Associate for multi document projects, plus Playbooks for codifying reusable negotiation standards, Spellbook in-house counsel review 2026-07-22 | Full |
| Triggers & Channel CoverageAgents are invoked from within the Word add in against a document or document set, but channel coverage beyond Word and event or schedule based triggers were not documented, Spellbook product pages 2026-07-22 | Partial |
| Model Flexibility & RoutingSpellbook is built primarily on OpenAI's GPT family with additional models from Anthropic and others plus its own fine tuning and prompting layers, indicating multi provider routing, but customer facing model selection was not documented, Spellbook comparison analysis 2026-06-11 | Partial |
| APIs, SDKs & MCP ExtensibilityPlaybooks allow configuration of agent behavior but no API, SDK, MCP surface, or customer agent builder was retrieved in this pass, Spellbook product pages 2026-07-22 | Unable to verify |
| Testing, Debugging & OptimizationSpellbook states Associate checks its own work and adapts as a project progresses, and users can modify agent behavior mid project, but agent testing, simulation, or versioning tooling was not documented, Spellbook Associate launch coverage 2024-08-22 | Partial |
| Browser & Computer UseNo browser or computer use capability was documented, Spellbook product pages 2026-07-22 | Unable to verify |
The Agentic Index coverage score grades every vendor Full, Partial or Unable to verify against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Recent platform changes
Spellbook launched an AI Document Editor within its Associate platform, allowing lawyers to review, edit, and finalize contracts natively without switching to Microsoft Word. The editor enables users to prompt the AI to make targeted, multi-document redlines, while retaining the ability to manually amend text, add comments, and accept or reject AI suggestions in a unified interface.
Bears on: Workflow orchestration
View sourceSpellbook launched a native integration for Google Docs, bringing its AI contract review, playbooks, and Ask Q&A features directly into the browser-based document editor. The integration allows legal teams to apply redlines and run reviews directly on the page. This bypasses the previous requirement to download documents and open them in Microsoft Word.
Bears on: Integrations
View sourcePricing
$99/user/mo
seats
Included quota
Entry seat (~$99/user/mo) bundles in-Word AI clause suggestions, drafting, and risk flagging; higher tiers add Review mode, custom clause libraries/playbooks, benchmarking, and the Associate agent.
What is public
Spellbook's official page states pricing is custom, per-seat, based on the number of licensed team members and set via a demo — no fixed list. A 7-day free trial is offered (extended trials for larger orgs), and academic institutions get free access. Triangulated figures: entry ~$99/user/mo, professional/team ~$149, enterprise ~$199–$350+/user/mo.
Billing mechanics
Per-seat annual licensing tiered by feature set (drafting/redlining vs. advanced playbooks, benchmarking, and the Associate multi-document agent) and contract length; annual billing runs ~20% under monthly. Enterprise plans (10-seat minimum) carry a 6-month minimum commitment after a late-2025 hike to ~$350/user/mo. Requires a Microsoft Word license.
Cost watchouts
Microsoft 365/Word required, 10-seat minimum + 6-month commitment on enterprise, ~20% monthly surcharge vs annual, renewal uplifts (10–25%/yr reported), optional paid Playbook Build Service
Variable cost rationale
Flat per-seat licensing with no metered usage, so spend scales predictably with seats and tier — low variable exposure
Additional watchouts
Word-only add-in (no Google Docs); pricing is demo-gated; enterprise hike + 6-month lock-in in late 2025
Overage / add-ons
No metered usage overage on the seat model; cost grows by adding seats or moving to higher feature tiers (one third-party listing reports a monthly draft cap on the entry tier, unconfirmed).
Sales call required
Yes, required for paid access
Lowest paid plan
From $99/user/mo; Pro tiers ~$329-399/user/mo
Commercial notes
~20% annual discount; volume discounts for 5–10+ seats; free academic access; built by Rally Legal; SOC 2 Type II / HIPAA / GDPR / EU AI Act
Key ambiguities
No official tier list or per-tier feature breakdown is published; entry/pro/enterprise dollar figures are triangulated and a reported entry-tier monthly draft cap is unconfirmed
Cancellation / refund
Custom annual contracts; enterprise carries a 6-month minimum; a published Refund Policy exists; renewal increases reported at 10–25%/yr without a cap
Support SLA / resale
Email/standard support at entry; priority support, dedicated support, group training, and custom playbook building on team/enterprise
Missing data
Spellbook does not publish fixed pricing; all figures are triangulated user/analyst reports, not list prices. Exact tiers, per-tier features, seat minimums, and the enterprise rate are quote-specific.
Related vendors
- 11th Estate — Agentic platform whose AI engine scans markets, matches an…
- Aera Technology — Decision intelligence platform where an always on agent executes…
- Akro AI — On premise operational intelligence that automates document heavy…
- alfred_ — AI executive assistant that triages the inbox overnight and scores…
- Algebra AI — Delivery led provider that builds and runs human governed AI agents…
- Alloy.ai — Commerce intelligence system for consumer brands that unifies four…
Alternatives to Spellbook
The closest documented capability profiles to Spellbook among enterprise operations agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Firmbase6.5 / 14Adds documented Observability & Auditability and Memory & State Persistence
- Mandel AI4.5 / 14Adds documented Observability & Auditability
- Pando5.5 / 14Fuller documented coverage on Human Oversight & Guardrails and Triggers & Channel Coverage
- Brightwave4.0 / 14Adds documented Observability & Auditability
- DataMission.ai4.0 / 14Adds documented Observability & Auditability
- effie.ai4.0 / 14Adds documented Observability & Auditability
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded