About Rephrasa

Who builds Rephrasa

Rephrasa is built and run by Ahmad Thariq Syauqi through Creativism, a digital marketing agency based in Yogyakarta, Indonesia. It is a small, independent project: there is no investor, no team of named experts and no partnership with any university or style guide publisher. When a page explains a citation rule, it links the official guide or library guide it relies on, so you can check it yourself.

The site has two editions. The English edition at the root of rephrasa.com serves students everywhere; the Indonesian edition under /id is written for students who write in Bahasa Indonesia. Both use the same tools.

What Rephrasa does, and what it does not do

• The paraphrasing tool rewrites text in Academic, Simple or Formal mode and keeps citations, numbers and technical terms. Free use allows 500 words per request and 2,000 words per day.
• The citation generator formats references in APA, MLA, Chicago, Harvard, IEEE and AMA from a DOI, URL or ISBN, free and unlimited.
• Rephrasa does not offer an AI detector, a plagiarism checker or Turnitin access, and it is not marketed as a way to get past any of them.

Privacy by design

Your text is not stored. When you paraphrase, your text is sent once from our server to the AI provider and the answer is sent back to you. Rephrasa does not write the text or the answer to logs, files, a database or a cache. Our logs hold only technical facts such as the mode, the language, the word count, the model and the response time. The tool never saves your text in your browser either.

The AI providers. The primary model is OpenAI gpt-6-luna; if it fails, Anthropic Claude Haiku 5.5 is asked once. Both are used through their business APIs:

• OpenAI states that "data sent to the OpenAI API is not used to train or improve OpenAI models" and that abuse-monitoring logs are "retained for up to 30 days" (OpenAI, Your data, checked October 25, 2026).
• Anthropic states: "By default, we will not use your inputs or outputs from our commercial products" for model training (Anthropic Privacy Center, checked October 25, 2026), and that it deletes API inputs and outputs "within 30 days of receipt or generation", keeping them up to 2 years only when a request is flagged as violating its usage policy (Anthropic data retention, checked October 26, 2026).

So a provider may keep a request for up to 30 days under its own policy (longer only for a request flagged as abuse), even though Rephrasa keeps nothing. Do not paste passwords, health records or other data you would not send to an online service.

Usage counting. To enforce the free daily limit, we keep a keyed hash of your connection and browser for the current day, plus a random cookie id, with word counters. Raw IP addresses and browser strings are not stored, and counters older than the previous day are deleted.

Citations stay on your device. Your bibliography is saved in your browser's local storage and never uploaded. DOI and ISBN lookups go from your browser to Crossref, doi.org and Open Library; URL lookups pass through our server, which reads only the page's metadata.

Analytics. We use Google Analytics with consent mode to count page views and button clicks. The text you paste and the sources you cite are never sent to analytics.

Usage rules

• The AI paraphraser is for people aged 18 and over. The tool asks you to confirm this the first time you use it.
• Paraphrasing a source does not make its ideas yours. Cite every source you use, and follow your institution's rules on AI writing tools.
• Requests to get past AI detectors, plagiarism checkers or Turnitin are refused. Rephrasa does not sell or provide Turnitin access.
• Automated or bulk use that gets around the security check or the daily limits is not allowed.

Open sources and licences

Citation styles and locales come unchanged from the Citation Style Language project (styles and locales repositories), licensed under CC BY-SA 3.0. Each style file names its own authors. The full attribution list is in LICENSE-and-ATTRIBUTION.txt.

Citations are formatted by citeproc-js, available under the CPAL 1.0 or AGPL licence; its source code is at that link and Rephrasa uses it without changes. Book data comes from Open Library, article data from Crossref and DOI content negotiation.

The interface fonts, Poppins and Lexend, are used under the SIL Open Font License.