Frequently asked questions

Every search runs in parallel against four engines: DuckDuckGo (general web), OpenAlex (250M scholarly works), Semantic Scholar (peer-reviewed papers), and CORE (200M open-access). We deduplicate URLs across engines so the same paper never wastes a slot. Academic engines also return rich metadata — authors, year, DOI, citation count — which appears in your results.

Yes. ProSearch detects your query's language and finds matching topics across languages. If you've previously searched a topic in English and now ask the same thing in Arabic, you'll see your existing library — not a fresh empty topic. The interface itself toggles between Arabic (RTL) and English from the menu.

One credit ($3.00) buys one research request: discovery + crawl + extraction across all four engines, with up to 60 sources stored permanently in your library. Once paid, browsing your sources, asking the AI assistant questions, generating reports, and exporting to BibTeX is free — no per-action fees.

Your credit is refunded automatically. If the platform is restarted while your search is in flight, the in-progress job is detected on the next boot, marked as failed, and the credit lands back in your balance — no support ticket needed.

Open any topic and click 'Ask AI'. Type a question — in any language. We embed it, retrieve the most relevant pages from your library using a multi-angle search (general / methodology / limitations / gaps / recent), then feed that focused context to an advanced AI model. Every claim in the answer cites a specific source you can click to verify.

Yes. From any results page, the BibTeX and RIS buttons download a ready-to-import file. BibTeX works with Zotero, Mendeley, JabRef, and any LaTeX project. RIS works with EndNote and RefWorks. Each entry includes authors, year, DOI, journal, citation count when available.

Your account, search history, chat conversations, and credit ledger are private to you. Topics are deduplicated across users (so the same Wikipedia article isn't re-crawled by every researcher), but your specific queries and AI conversations are never shared or used to train models.

There's no hard cap on your library. Each search adds new sources and skips ones you already have. A topic that started with 30 sources can grow to thousands as you re-search it over time — and the AI assistant works equally well at every scale because it retrieves the most relevant subset for each question.

Click your name in the top-right corner, then 'Settings'. You can change your password (current password required) and set your preferred language. The language menu in the dropdown switches the UI for the current session.