SCORE THE FINE PRINT

About 2terms

What Is 2terms?

2terms is a scorer for consumer terms of service and privacy policies that runs entirely in your web browser. It reads a policy, finds the clauses that matter to you, quotes each one word for word with its location, and turns them into nine category scores and one overall score from 0 to 100.

The method is deterministic and published: the same text always produces the same score, and every pattern it looks for is listed on the methodology page with a plain-English meaning.

How Does the Scoring Work?

  1. Upload the policy as a PDF with selectable text, a Word .docx, a saved .html page or a .txt file, or paste the text.
  2. The document-type check confirms it is a terms of service, privacy policy, cookie policy, consumer license or AI terms document, and turns anything else away.
  3. The text is split into sections and clauses, keeping each clause with the heading it sits under.
  4. Every clause is compared with the published pattern library: 74 patterns across nine categories, each marked as raising or lowering the impact on you.
  5. Each category the document addresses starts at 50 and moves by the points of each distinct pattern found, at most 25 per pattern; categories it does not address are not graded. The overall score is the weighted average of the graded categories, shown with a letter band from A to E and how many of the nine categories were used.
  6. You read every finding with the exact clause quoted and its location, and can follow each one to its entry on the methodology page.

What Do the Nine Categories Measure?

Each category is scored from 0 to 100 and weighted by how much it can affect you. The weights add up to 100 and come from the 2terms research verdict.

  • Data collection & surveillance (weight 18). What personal data the document says is collected, from where, and how invasive it is (precise location, biometrics, tracking across other sites and apps, data bought from others).
  • Third-party sharing & sale (weight 16). Whether personal data is sold, shared with advertisers or data brokers, or kept within the company and its service providers.
  • AI training on user data (weight 14). Whether your content, prompts or data may be used to train or improve AI or machine-learning models, read by human reviewers, or passed to third-party model providers, and whether you can opt out.
  • Legal rights (weight 14). How disputes are handled: mandatory arbitration, class-action and jury waivers, liability caps, short claim deadlines, and protections such as an arbitration opt-out or small-claims option.
  • User-content license (weight 10). What rights you grant over the things you post or upload: scope, duration, sublicensing, use of your name or likeness, and whether you keep ownership.
  • Retention & deletion (weight 10). How long data is kept, whether it survives account deletion, and whether you can delete or export it within a stated time.
  • Security commitments (weight 8). Concrete security commitments (encryption, end-to-end encryption, breach notification, audits) versus disclaimers of responsibility.
  • Changes & termination (weight 6). Whether the terms can change without notice, whether continued use counts as acceptance, and whether your account can be ended for any reason.
  • Children's data (weight 4). Whether the service is aimed at children, how it treats data of users under 13 (or under 16/18), parental consent, and advertising to minors.

Which Documents Can You Score?

2terms accepts the documents a consumer agrees to when using an app or website:

  • Terms of Service / Terms of Use
  • Privacy Policy / Privacy Notice
  • Cookie Policy
  • End-user license agreement (consumer software)
  • AI / Generative-AI terms

Negotiated or signed documents are turned away with an explanation: contracts, leases, non-disclosure agreements, employment agreements, data processing agreements, court filings and invoices. The check looks for markers such as signature blocks, "WITNESSETH" and "IN WITNESS WHEREOF".

How Is Your Document Kept Private?

The analyzer is a static web page. When it opens, your browser downloads the page's code, its fonts and the PDF reader. After that, reading and scoring a document needs nothing from any server.

  1. Open your browser's developer tools with F12, or Cmd+Option+I on a Mac, and select the Network tab.
  2. Clear the list, then upload a file or paste text and score it.
  3. The list stays empty: the analysis made no request. The counter beside the analyzer shows the same thing, read from your browser's own Resource Timing records.

Document text is never written to local storage, cookies or any other browser storage. The only thing 2terms stores is your light or dark theme choice.

Where Do the Pre-Scored Services Come From?

The service pages were built offline by running the same scorer on archived copies of each service's terms and privacy policy. Current and recent versions come from Open Terms Archive, which records published terms over time and releases them under the Open Database License. Older versions of ten headline services come from Internet Archive Wayback Machine snapshots, one per year from 2016.

Every version on a service page names its source and date. Where an archive has no usable copy for a year, the year is simply missing; nothing is filled in.

Who Is 2terms For?

  • People deciding whether to sign up for an app and wanting to know what they agree to.
  • Parents checking how a service treats children's data.
  • Journalists and researchers comparing policies, with every claim backed by a quote.
  • Small businesses checking the terms of tools they ask staff to use.
  • Anyone curious how a policy changed over the years.

What Are the Limits of a Pattern-Based Score?

Patterns read sentences, not intent. A clause can be worded in a way no pattern expects, and a pattern can occasionally match a sentence that means something narrower. Categories a document does not mention are shown as not addressed and left out of the overall score, and every result says how many of the nine categories were graded, so a short policy is not automatically a safe one. Version 1 reads English only and does not do OCR. The full list of known limitations is on the methodology page.

Frequently Asked Questions

Does my document leave my device?

No. The file or text you give 2terms is read and scored by code that is already loaded in your browser tab. It is never uploaded, never stored in the browser, and it is gone when you close the tab. The site also sends a security policy that tells your browser to refuse connections to any other website.

Is 2terms legal advice?

No. 2terms quotes the clauses it finds and scores them with a published, deterministic method that describes how each clause affects the user. It does not say whether a clause or a company complies with any law. For advice about your situation, talk to a lawyer.

Does 2terms use AI to read my document?

No. Scoring uses a fixed library of text patterns that is published in full on the methodology page. No AI model or outside service is involved, so the same text always gets the same score.

Which documents can 2terms score?

Terms of service and terms of use, privacy policies and notices, cookie policies, consumer end-user license agreements, and AI or generative-AI terms, in English. Contracts, leases, NDAs, employment agreements, data processing agreements, court filings and invoices are turned away with an explanation.

Why was my PDF rejected?

A PDF is rejected when it has almost no selectable text, which usually means it is a scan or a photo. 2terms version 1 does not do OCR. Open the policy in your browser and paste its text instead. Files over 10 MB are also rejected.

What does a high score mean?

Scores run from 0 to 100, where 0 means the matched clauses have the least impact on you and 100 the most. Each category the document addresses starts at 50; clauses that expand what the company can do with your data or rights raise it, and commitments that protect you lower it. No single clause can move a category more than 25 points, so extreme scores require several separate findings. Categories the document does not address are not graded, and the result says how many of the nine were used.

Where do the pre-scored services come from?

The pre-scored pages run the same scorer offline on archived copies of each policy: versions recorded by Open Terms Archive (published under the Open Database License) and, for older years, snapshots from the Internet Archive Wayback Machine. Every page names the source and date of each version.

How often are the service scores updated?

Version 1 scores were built on November 19, 2026 from the Open Terms Archive release of November 16, 2026. Each service page shows the date of the latest version that was scored.