Plain-language literature search
Ask research questions conversationally instead of writing query syntax or Boolean filters. Curie searches PubMed, arXiv, Europe PMC, OpenAlex, and Semantic Scholar in parallel.
Curie
Freemium
Curie is an AI research assistant for searching scientific literature, analyzing uploaded documents, extracting evidence, and organizing references. It is designed for solo researchers, graduate students, systematic review teams, and R&D groups.
Curieとは？
Curie is an AI research assistant for scientific literature and document analysis. It accepts research questions in plain language, searches five scientific databases in parallel, analyzes available full text, and checks answer claims against the source passages behind them.
The product combines literature search with document upload, systematic review support, a saved research library, and structured data extraction. It is intended for solo researchers, graduate students, systematic review teams, and industry R&D groups that need to search, screen, summarize, compare, or extract evidence from scientific documents.
Curieでできること
Claim verification
Curie checks claims in its answers against the source they came from and indicates when a claim cannot be verified. When full text is available, it can use the paper rather than only its abstract.
Document analysis and uploads
Upload PDF, DOC/DOCX, TXT, or Markdown files and analyze them alongside searched literature. The FAQ specifies a 50MB per-file limit and up to 20 files or 300MB in one upload.
Systematic review workflows
Starter and higher plans support full systematic reviews, including a PRISMA diagram. Curie can also help with screening, evidence extraction, summaries, and review-protocol drafting.
Library with tags, notes, and exports
Save articles and uploaded documents in one library, add tags and autosaved notes, filter by source or year, and export selected references as BibTeX, RIS, or CSV.
Quote-backed data extraction
Define columns such as text, number, yes/no, or fixed categories for one document or saved article. Curie fills the table with an exact supporting quote behind each value and can export the result as CSV.
利用シーン
“Starting a literature review”
Ask a broad research question in natural language, review ranked results and source-backed answers, then save relevant articles into a tagged reading list for later screening.
“Building a systematic review”
Use literature search, document uploads, evidence extraction, and review support to assemble a structured workflow. Starter and higher plans include full systematic reviews and a PRISMA diagram.
“Standardizing study data”
Define a consistent set of columns—such as author, year, sample size, dose, or primary outcome—and run individual papers through Extract Data to produce quote-backed rows for a spreadsheet.
“Handing off references to collaborators”
Filter a library by project, theme, source, or year and export the selected references as BibTeX, RIS, or CSV for a reference manager or shared research workflow.
“Checking a scientific claim”
Ask whether a claim is supported by the literature and inspect the source quote attached to the answer. If the claim cannot be verified, Curie reports that rather than leaving it unflagged.
よくある質問
Do I need special search syntax or training to use Curie?
No. Curie is designed for plain-language questions and does not require query syntax or setup training.
Which scientific sources does Curie search?
Curie searches PubMed, arXiv, Europe PMC, OpenAlex, and Semantic Scholar in parallel for each question.
What file types and upload limits are supported?
Supported formats are PDF, Word (.doc and .docx), plain text, and Markdown. The stated limits are 50MB per file and up to 20 files or 300MB total in one upload.
Can I export results to a reference manager or spreadsheet?
Yes. Library selections can be exported as BibTeX, RIS, or CSV. Extract Data results can be downloaded as CSV.
What happens when I reach my plan’s usage limit?
Each plan has a monthly usage budget. When it is exhausted, further paid actions such as starting a new analysis pause until the next billing cycle.
クイック情報
- Category
- AI research assistant; scientific literature and document analysis
- Primary users
- Solo researchers, graduate students, systematic review teams, and R&D departments
- Literature sources
- PubMed, arXiv, Europe PMC, OpenAlex, and Semantic Scholar
- Plans
- Free; Starter at $19/month plus VAT; Pro at $49/month plus VAT; Team pricing by request
- Reference exports
- BibTeX, RIS, and CSV
- Data and hosting
- GDPR compliant; EU-hosted infrastructure; user data not used to train or improve models
Curieの代替品
WisPaper
wispaper.ai
WisPaperは、科学者や研究者向けのAI研究ワークスペース。論文検索、選別、レビュー、文献管理を支援します。
Rayyan
rayyan.ai
系統的レビューと文献レビューのためのAIプラットフォーム
Anara
unriddle.ai
Anaraは、文献・ウェブ検索、出典付き回答、文書分析を一元化するAIリサーチワークスペースです。
Windward
windward.ai
船舶審査・監視・コンプライアンス向け海事 AI
Termo
termo.ai
Termoは各エージェントが専用VMで動く、調査・自動化・開発向けAIエージェント基盤です。
ResearchGPT
researchgpt.com
AI支援の株式調査。ティッカーや質問から出典付き投資メモを作成。