Unriddle is positioned as an AI research assistant and workspace for scientific research, designed to help researchers complete tasks such as literature search, library management, cross-document Q&A, and writing with citations, emphasizing the verifiability of "showing the work process."
Purpose The tool is mainly used for research literature-related workflows, including finding research papers across fields, asking questions within one's own library, synthesizing large numbers of documents to extract key insights, and automatically generating verifiable citations in writing. It is officially claimed to be 4 times more accurate than general-purpose AI, with every answer pointing to the exact passage of its source and verifiable in one click.
Main capabilities
- Literature search: Supports searching over 300M papers, filtering by journal and year, sorting by citation count or recency, viewing reference lists, displaying PDF availability, and filtering by criteria. Covered fields include medicine, physics, and others.
- Library building: Allows uploading various file types such as papers, notes, slides, and recordings, supports importing via DOI, PMID, or URL, can sync with Zotero and Mendeley, supports OCR recognition for scanned PDFs, automatically fills metadata, and supports folder and tag management.
- Cross-source Q&A: Allows asking questions about a single paper, a folder, or the entire library, provides paragraph-level citations in answers, can be limited to answering only based on the library, and supports highlighted questions, comparing papers, setting folder instructions, and more.
- Table extraction: Can turn folder contents into a table, with one row per paper and one column per question, all cells with citations, supports defining columns, numeric and date columns, syncs with folders, and can export CSV and Excel.
- Writing and citation: Can draft in notes and cite suggestions from the library, supports any citation style or custom styles, generates reference lists and BibTeX, can export to Word, LaTeX, or BibTeX, and provides a Microsoft Word plugin and AI quick editing features.
- API: Provides the Anara API for searching 300M+ papers.
How to use After users upload or import literature materials, they can search, ask questions, and compare papers within the library, and output results in the form of cited tables or notes. Citations during writing come from the library, and export to multiple formats is supported. The official offering also includes iOS (iPhone, iPad) and Microsoft Word integration, as well as a PDF reader.
Use cases Officially showcased use cases include: a medical technology R&D team using it for literature indexing, searching, and summarization for clinical research protocols and FDA regulatory submissions;a life sciences consulting team completing a systematic review across thousands of papers within a short cycle; a doctoral student asking questions about 700 papers collected over many years and discovering previously unnoticed themes; a biotechnology scientist quickly finding papers; a regulatory expert obtaining feedback that identifies gaps; and an academic researcher verifying information by highlighting text.
Limitations and notes Official materials do not provide specific pricing, free quota, or paid plan details, so they are not available for now. They also do not explain details such as the number of supported languages or specific file size or quantity limits, so these are not available for now. Regarding security, the official statement says it is certified for SOC 2 II and ISO 27001 and compliant with GDPR and HIPAA. The tool is backed by Y Combinator, and the official notice marks © 2026 Anara Labs, Inc.
