Lumina

LuminaAI literature review software

Lumina is an AI tool for systematic reviews and literature screening. It connects to multiple academic databases, removes duplicates, and prioritizes records according to the inclusion criteria set by researchers. Human reviewers retain the final decision, while audit records track the screening process. The platform also supports two independent reviewers, in-app conflict resolution, open-access PDF retrieval, custom PDF uploads, and PRISMA 2020 flowchart exports. Lumina offers a five-day free trial for one project with up to 75 records. Full workflow access requires a subscription, but the website does not publish pricing. Its AI ranking is designed to assist screening rather than replace researcher judgment.

freemium
systematic reviewliterature screeningAI research toolsPRISMAacademic database searchdual reviewer screeningAI-assisted evidence synthesis
Indexed
0.0 (0 Number of reviews)

Log in to rate the project

Try Now

Lumina is an AI tool for systematic reviews, positioned by its website as “AI Literature Review & Systematic Review Software.” Its central approach is to rank records so that the papers most likely to match a review appear earlier, while researchers retain responsibility for inclusion decisions.

Search, import, and screening

The platform brings several parts of a review workflow into one project. Researchers can search Scopus, PubMed, OpenAlex, Europe PMC, arXiv, bioRxiv/medRxiv, and OpenAIRE, with automatic deduplication. Existing libraries can also be imported in RIS, CSV, or BibTeX format, including files from tools such as EndNote and Zotero.

After records are imported, Lumina ranks abstracts according to the user’s inclusion criteria. Each decision is recorded for auditing, but the AI only prioritizes the queue; it does not make the final inclusion decision. The website claims users can screen 5,000 papers in one evening. That is an official marketing claim, and actual results will depend on the review context.

Collaboration and reporting

  • Dual screening: Two reviewers can screen independently and resolve disagreements inside the application.
  • Full-text access: The platform can retrieve open-access PDFs and accepts user-uploaded PDFs.
  • PRISMA export: A built-in PRISMA 2020 flowchart generator supports structured export.
  • Traceability: The workflow covers importing, deduplication, AI-assisted screening, full-text retrieval, conflict resolution, and export within a single project, with actions kept traceable.

Trial and subscription

Lumina offers a free trial limited to one review project with up to 75 records. The trial lasts five days and starts only after the project has been successfully created. Decisions can still be exported during the trial, but the complete workflow requires a subscription. Lumina’s website does not disclose subscription pricing.

For teams conducting systematic reviews or meta-analyses, the practical benefit is having database search, deduplication, screening, collaboration, and PRISMA reporting in one interface. Researchers still need to assess each paper themselves, and the company has not published accuracy or benchmark data for its AI ranking.

Method: This page was prepared with AI assistance from the official site or repository, with key numbers automatically checked against that source. Features and pricing may change; verify them at the source. official product page

Pros & Cons

Pros

  • Combines multi-database search and automatic deduplication
  • Uses AI to prioritize screening while preserving human final decisions and audit trails
  • Includes a PRISMA 2020 flowchart generator
  • Supports independent dual screening and in-app conflict resolution

Cons

  • Free trial is limited to 75 records and five days
  • Full workflow requires a subscription, with pricing not publicly disclosed
  • No published accuracy or benchmark data for the AI ranking

Frequently Asked Questions

Which literature import formats does Lumina support?

Lumina supports RIS, CSV, and BibTeX files. Researchers can also bring in libraries from tools such as EndNote and Zotero, provided the records are exported in a supported format. This makes it possible to start with an existing reference collection rather than conducting every search inside Lumina.

Which academic databases can Lumina search?

The website lists seven sources: Scopus, PubMed, OpenAlex, Europe PMC, arXiv, bioRxiv/medRxiv, and OpenAIRE. Lumina also supports automatic deduplication when records are brought together, which is useful when a review draws results from multiple databases.

Does Lumina support collaboration between reviewers?

Yes. Lumina supports two reviewers working independently on the same screening process. When their decisions differ, the reviewers can resolve the conflict within the application. The platform also keeps an audit trail of decisions, allowing the review process to remain traceable.

What are the limits of Lumina’s free trial?

The free trial is limited to one project containing no more than 75 records and lasts five days. The countdown begins after the project has been successfully created, rather than when a user first signs up. Decisions can be exported during the trial, while the complete workflow requires a subscription.

Explore More

Open-source Alternatives

awesome-ai-research-writing: AI Paper Writing Resources

awesome-ai-research-writing is a GitHub collection focused on AI research writing. It brings together tools, templates, practical techniques, and related reading intended to reduce the repetitive work behind drafting, revising, and polishing papers or technical reports. With more than 33,000 GitHub stars at the time of review, the repository has attracted substantial community attention. Its main value is not that it replaces an author or supervisor, but that it gives researchers a single place to begin looking for useful writing resources. Students, research engineers, and academic writers can browse the README, identify relevant entries, and test them against their own workflow.

Awesome AI for Science: Curated AI Resources for Scientific Discovery

This GitHub repository offers a curated list of AI tools, libraries, papers, datasets, and frameworks spanning physics, chemistry, biology, and materials science. It serves as a valuable resource for researchers and developers to quickly grasp and apply AI in scientific exploration, with over 1,700 stars and an MIT license.

earth2studio: NVIDIA Deep Learning Framework for Weather and Climate

earth2studio is an open-source deep learning framework from NVIDIA, designed for the weather and climate domain. It streamlines the workflow from research to deployment, offering universal APIs and pre-trained models. This enables researchers to rapidly develop AI-driven weather forecasting and climate simulation applications, lowering barriers and accelerating innovation in the field.

ai4paper: Open-Source AI Platform for Researchers

ai4paper is an open-source AI platform designed for researchers, claiming access to 240 million academic papers. Core features include full-text PDF translation, AI-driven literature search, and one-click review generation, all accessible via a web interface without plugins. It offers Zotero integration and journal subscription via mini-programs, aiming to boost efficiency in literature review and academic writing. The project is primarily written in HTML, licensed under MIT, and had 2739 stars on GitHub at the time of collection.

openscience: An Open-Source AI Workbench for Research

openscience is an open-source AI workbench from synthetic-sciences, specifically designed for scientific research. Built with TypeScript, the project has garnered over 3.2k stars on GitHub, featuring a comprehensive repository with frontend, backend, CLI, and evaluation modules. While public documentation is currently limited, it's a project worth watching for teams interested in AI for Science.

ResearchStudio: Microsoft Open Source AI Collaboration Tool

ResearchStudio is an open-source AI collaboration tool from Microsoft, designed to support researchers through the entire academic journey from initial problem formulation to final publication. It integrates features for literature review, experimental design, data analysis, and paper writing, leveraging large language models to provide intelligent suggestions. The project is particularly suited for academic researchers seeking to streamline their workflow. The primary language is Python, the license is MIT, and it had 1911 GitHub stars at the time of collection.