Cubby Law logo
Education · Learning Growth

Cubby Law

Cubby Law is an AI teaching assistant for law school. It learns how your professor teaches, tests and grades, then generates professor-specific practice exams, quizzes and case briefs with instant, rubric-based feedback, backed by a top-half refund guarantee.

Active Free trial Subscription No public API 13+ Verified by Guidaio
Overview

What is Cubby Law?

Cubby Law is an AI teaching assistant built for one audience: students in American law schools, where a single final exam usually decides the grade and the curve decides everything else. Its premise is that a general-purpose model is the wrong instrument for that problem. ChatGPT knows the law, but it does not know that your professor tests a minority rule, weights an IRAC answer in a particular way, or returns to the same fact patterns every year. Cubby is tuned to exactly that.

Students upload an outline, class notes, slides and past exams, and the platform adapts to the professor's emphasis and exam style. The company also claims to hold prior analysis of real course materials from many professors, so a new account does not start from nothing. From that corpus it produces unlimited quizzes, professor-specific practice exams and instant rubric-based grading that shows where points were earned and lost. More than 40,000 case briefs are available. Questions go to Donna, a closed-system assistant whose answers are grounded in the uploaded materials and cite back to them, a deliberate defence against the hallucination the company attributes to open-web models.

A separate product serves faculty. Professors upload a syllabus and, in roughly twenty minutes, obtain generated assessments, drafted rubric feedback and a dashboard showing class averages, completion rates and the topics a cohort is missing. Nothing reaches a student without the professor's approval.

The company publishes its own evidence. In a blind evaluation of 1,000 matchups in July 2026, law students preferred Cubby's answers 63% of the time, against 44% for ChatGPT, 36% for Claude and 29% for Gemini. A survey of 200 students after the autumn 2025 semester reported 90% finishing in the top half of their class and 55% in the top fifth. Cubby backs this with a full refund for annual subscribers who finish below the median, under published and fairly demanding conditions.

It is web-only: no mobile application, no public API, no live LMS integration. Faculty advisers at Brooklyn, Quinnipiac, Emory and Texas are named on the site.

What it does

  • Generate unlimited practice exams written in the style of your own professor
  • Grade answers instantly against a rubric, showing where points are earned and lost
  • Build quizzes from the outline, notes and slides you upload
  • Consult more than 40,000 case briefs annotated with your professor's angle
  • Answer doctrinal questions through Donna, a closed-system assistant that cites your materials
  • Draft professor-quality multiple-choice questions and fact patterns from a course corpus
  • Show faculty which topics a class is failing and which students have stopped working
Audience

When to use Cubby Law / When not to

A quick filter to help you decide if Cubby Law is the right fit.

When to use Cubby Law

  • Law students graded on a curve, where finishing above the median shapes the first salary
  • 1Ls sitting their first closed-book law school exams with no idea how they are being marked
  • Law professors who want assessments and rubric feedback drafted from their own course materials
  • Academic support staff and exam-prep tutors coaching law students through a semester
  • Directors of academic excellence who need to see which topics a cohort is failing in October

When not to use Cubby Law

  • Students outside American law school: the entire product is built around US JD coursework
  • Practising lawyers looking for legal research, drafting or matter management tools
  • Anyone who needs a mobile app, a public API or a browser extension, as Cubby is web-only
  • Schools that require a live LMS integration, since Canvas and its equivalents are not connected
  • Students whose honour code forbids uploading professor materials to third-party platforms
Get started

How to use Cubby Law

A typical end-to-end flow, from setup to results.

  1. Sign up from the homepage using the Get Started button
  2. Create your courses; the site claims setup takes under ten minutes
  3. Upload your outline, class notes, slides and any past exams you hold
  4. Let Cubby process the material and organise the course around the topics it finds
  5. Ask Donna about the doctrines you have not yet pinned down
  6. Generate quizzes targeting those specific doctrines
  7. Sit a full practice exam written in your professor's style
  8. Read the instant rubric-based grade to see where points were won and lost
  9. Use the case briefs to prepare for class and cold calls
  10. Professors: review and approve every exam, score and comment before students see it
Quick read

Pros & Cons

Pros

  • Genuine specialisation: the tool is calibrated on a course and a professor, not merely on a subject
  • Published performance figures with a described methodology, including a 1,000-matchup blind evaluation
  • A money-back guarantee tied to class rank, with its conditions published in full
  • An explicit data commitment to faculty: no general model training, no pooling, encryption in transit and at rest
  • A closed system whose answers cite back to the user's own corpus
  • Professors keep publication control: no draft grade or comment reaches a student unapproved
  • Low price against the legal study market, at 300 USD a year

Cons

  • No permanent free plan for students: the commitment is paid from day one
  • Very narrow scope, confined to American law school coursework
  • No live LMS integration; Canvas and equivalents are only on the roadmap
  • Web-only, with no mobile application and no public API
  • Every performance figure is self-reported by the vendor, with no independent verification
  • A generic, dated privacy policy that states data may be kept indefinitely and permits sharing with marketing partners
  • The top-half guarantee is tightly conditioned: annual plan, purchase window, usage thresholds and official proof of rank
Pricing

Pricing & Plans

There is no permanent free plan for students. The cheapest paid entry point is 25 USD per month, billed annually at 300 USD, which the vendor presents as a 29% saving; a month-to-month subscription costs 35 USD per month. Access is granted immediately on payment. Professors are offered a free pilot for themselves and their students.

Plan 1
  • Annual — 25 USD/month
  • billed 300 USD yearly
  • 29% saving
  • all courses all year
  • top-half guarantee included
Plan 3
  • Professor pilot — free to pilot
  • for the professor and their students
Plan 4
  • All paid plans include unlimited practice exams in the professor's style
  • instant rubric-based grading
  • quizzes generated from uploads
  • and 40
  • 000+ case briefs
Special offers — Top-half guarantee: a full refund of fees if the student does not finish in the top half of their class, reserved for annual subscriptions bought between 15 June and 1 December 2026 and applying to the autumn 2026 semester only · 29% saving on the annual subscription compared with the month-to-month rate · Free pilot for professors, covering the professor and their students
Prices and plans listed above may evolve. Always check the official pricing page before subscribing.
Trust & Privacy

Data, GDPR & hosting

A consolidated view of how Cubby Law handles your data.

GDPR overview

Cubby never claims GDPR compliance in so many words, and the distance between its contract and its privacy policy is worth naming. The terms of use define EU Data Protection Laws as the legislation implementing Regulation (EU) 2016/679, borrow the GDPR's own definitions of personal data and processing, and incorporate a Data Processing Agreement by reference whenever Cubby processes personal data on a customer's behalf. That DPA is not published on the site and does not cover beta services. Cubby also undertakes to tell customers of any legal obligation preventing it from following their instructions. The privacy policy, by contrast, is a plainly American document: it addresses California residents, sets the age floor at 13, and says nothing about data subject rights, retention periods or transfer safeguards. No Article 27 representative and no data protection officer are named.

Who owns the data?

The agreement transfers no intellectual property in either direction: Cubby gains no rights in customer data, and customers gain none in the service. Uploaded material stays the customer's, with Cubby granted only the limited permission needed to deliver the service, extended to its subcontractors and sub-processors. Professor materials are never pooled into shared models, never used to improve another professor's course and never added to a general training set. Two carve-outs matter. Aggregate and anonymous usage data is owned outright by Cubby, usable during and after the contract, though never disclosed in a form that identifies a customer. Any feedback sent to Cubby is assigned to Cubby.

Reuse rights

Professors can export or delete their materials at any time without asking permission, lock and unlock them, and decide what students see and when; nothing reaches a student before the professor approves it. Students can view or delete their personal data from their account or by writing to team@cubby.nyc. Users may share content with other users or third parties, but the customer alone carries the consequences of doing so. The restrictions are firm: no reselling or leasing the service, no reverse engineering, no using it to build a competing product, and no publishing benchmarks or comparative analyses without Cubby's prior written consent. After termination, additional time to export stored data may be granted and may be charged at the going rate.

Data retention & training

Retention summary
The privacy policy, last updated on 16 October 2025, is blunt: data may be kept indefinitely. No retention period is stated and no anonymisation schedule is published. Users can view or delete their personal data from their account or by emailing team@cubby.nyc, and professors can export or delete their materials at any time. On termination, Cubby deletes end user accounts and stored data within a commercially reasonable period, but only after an administrator requests it before the contract ends; extra time to export may be granted and may be charged. Aggregate and anonymous data derived from usage stays with Cubby afterwards. California residents may ask once a year what was shared for direct marketing.
Trains on customer data
No
DPA available
Yes
GDPR contact

Hosting summary

The site names no hosting country and no hosting region. The terms of use expressly allow Cubby and its subcontractors to transfer, access and store customer data in locations other than the customer's own country, which leaves the jurisdiction open. Security is described rather than certified: industry-standard technical and organisational measures, a security page referenced in the terms but not linked from the pages collected, and, in the commitment made to professors, encryption in transit and at rest, tightly restricted access and handling of student information with the care FERPA demands. No SOC 2 or ISO 27001 certification is published. Sub-processors are defined contractually and consented to by the customer, but no list of them is published; Cubby remains liable for their acts. If the security page is changed in a way that materially reduces protection, the customer is entitled to notice and may terminate within five business days. The site is served behind Cloudflare on an anycast address, which says nothing about where data actually rests.

Watch-outs

Things to keep in mind

Risks and trade-offs to weigh before adopting Cubby Law.

  • Honour code exposure: some law schools forbid uploading professor materials to any third-party tool, and the vendor places that responsibility squarely on the student
  • Cognitive dependence: drilling the exam format can quietly replace doctrinal understanding, a risk the company's own manifesto acknowledges without neutralising
  • Aggregate and anonymous usage data is owned by Cubby and remains exploitable after the contract ends
  • The privacy policy states that data may be kept indefinitely, with no retention period and no anonymisation schedule
  • The same policy permits sharing identifiable information with trusted partners for advertising and offers, and collecting precise location data
  • Liability is capped at the lesser of 1,000 USD or twelve months of fees, with mandatory individual arbitration and a class action waiver
  • For faculty, handing over past exams and model answers means entrusting a career's work to a young company with no published security certification
Setup

Setup & Integrations

Technical difficulty

Very low. Cubby is a web application and requires no installation, no server configuration and no API key. A student creates a course and uploads an outline, notes, slides and past exams; the site claims this takes under ten minutes. A professor needs about twenty minutes and can begin with a syllabus alone, adding exams, rubrics and slides later. Students join a course through an invitation link. Since no LMS integration exists, there is nothing to connect and nothing for an IT department to approve.

Deployment

Web app
Company

Behind Cubby Law

Company name
Cubby NYC, Inc.
Founded
27/02/2024
Country of origin
🇺🇸 United States
UBO
Truman Sacks
UBO country
🇺🇸 United States
Domain registrar country
🇺🇸 United States
Legal contact
Support contact

Fundraising

Seed round of 2.75 million USD, reported by the press at the time of the national launch in August 2025
Investors named: Eighty-Seven Capital (Sam Hinkie), Ludlow Ventures (Jonathon Triest), The Mini Fund (Eros Resmini, Discord co-founder), Positive Sum (Patrick O'Shaughnessy)
The site itself mentions no fundraising; the figures come from off-site press coverage and warrant human verification

Social

Official links

Resources

All the official URLs gathered for verification and reference.

Compare

Alternatives

Tools that compete with or complement Cubby Law.

C ChatGPTC ClaudeG GeminiQ QuimbeeB BarbriN NotebookLM
FAQ

Frequently asked questions

How is Cubby Law different from ChatGPT?
ChatGPT knows the law; Cubby is tuned to a specific class. Students are graded on a curve against the rules and frameworks their own professor teaches, and a general-purpose model will confidently return the majority rule even where the professor tests another one. Cubby trains on the professor's actual materials, past exams and model answers.
What does it cost, and is there a free plan?
There is no permanent free plan for students. An annual subscription costs 300 USD, presented as 25 USD per month, and a month-to-month subscription costs 35 USD per month. Professors are offered a free pilot for themselves and their students.
How does the top-half guarantee work?
Annual subscribers who use Cubby consistently through the autumn semester and do not finish in the top half of their class can claim a full refund. The published conditions are demanding: purchase between 15 June and 1 December 2026, minimum usage thresholds in exams and quizzes, at least fifteen separate active days, official documentation of class rank, and a request within thirty days of grades being released.
Are my uploaded materials shared with other users?
No. The vendor states that uploads stay private to the account and are never shared with other students, even in the same section or school. Professor materials are never pooled into shared models, never used to improve another professor's course and never added to a general training set.
Does using Cubby breach my school's honour code?
The vendor says no, on the ground that the tool does not complete assignments or produce submittable work. It also warns that some schools restrict uploading professor materials to any third-party platform, Cubby included, and places the responsibility on the student to check their own honour code.
Is there a mobile app or an API?
No. Cubby is a web application only. There is no iOS or Android app, no public API documentation, and no browser extension. Direct LMS integrations such as Canvas are described as not yet live.
What is the minimum age to use Cubby?
The terms of use set the floor at 13 for users residing in the United States and 16 for users residing anywhere else, or higher where local law requires a greater age for the service to be provided without parental consent.
How long does it take a professor to set up a course?
About twenty minutes, according to the vendor. Setup is mostly uploading material that already exists; a syllabus alone is enough to begin, and past exams, model answers, rubrics, slides and notes make the output closer to the professor's own standards.
Where is my data hosted, and how long is it kept?
The site names no hosting country or region, and the terms expressly allow Cubby and its subcontractors to store data outside the customer's country. The privacy policy, last updated on 16 October 2025, states that data may be kept indefinitely, while professors can export or delete their materials at any time.
Conclusion

Should you pick Cubby Law?

Cubby Law is unusually disciplined about what it is. Rather than another general assistant with a legal skin, it attacks one narrow and genuinely painful problem: American law school grades on a curve, one exam usually decides everything, and what counts is not the law in the abstract but the rules, emphases and marking habits of the professor in front of you. Building the product around that insight is what makes it defensible, and the comparison the company invites with ChatGPT is fair on its own terms.

The evidence is more encouraging than most of this market offers, and still self-reported. A blind evaluation of 1,000 matchups, a survey of 200 students, a 90% top-half figure and a refund guarantee are real commitments, but they were produced and published by the vendor, and the guarantee is hedged with a purchase window, usage thresholds and a documentation burden that will disqualify casual users.

The data story is split in two. The commitment made to professors is precise and reassuring: no general model training, no pooling across courses, encryption in transit and at rest, FERPA-aware handling, export and deletion on demand. The public privacy policy is the opposite, a generic and dated document that permits indefinite retention and sharing with marketing partners, names no hosting country and never addresses European rights. Anyone in the EU should read it before uploading anything.

The practical limits are clear: web-only, no API, no mobile app, no LMS integration, and no free plan for students. At 300 USD a year against a class rank worth six figures, the arithmetic is easy for the target user, and irrelevant to everyone else. This is a tool for one population, and it should be judged by them.