This directory represents an evidence synthesis ecosystem mapped in the associated systematic mapping study. It covers purpose-built evidence synthesis platforms plus general-purpose computational libraries and infrastructure utilities used to build and automate custom evidence synthesis workflows. It includes 291 verified open-source, non-proprietary tools available up to early 2026, spanning all stages of the pipeline. It is intended for systematic review researchers and students, meta-analysis practitioners, medical/public health/environmental researchers, librarians, methodologists, open science advocates, and research software developers.
The systematic process of identifying, evaluating, and integrating all available evidence relevant to a research question to inform science, policy, and practice - aiming for transparent, reproducible, and generalizable findings. Core components: systematic searching, screening and study selection, critical appraisal, data extraction, risk of bias assessment, and quantitative or qualitative synthesis.
The directory covers the full workflow: search strategy development, citation chasing and retrieval, reference management and deduplication, AI-assisted screening and prioritization, data extraction and PDF processing, risk of bias assessment and visualization, statistical analysis and meta-analysis, evidence mapping and bibliometrics, automation/scripting/machine learning, and qualitative synthesis.
Evidence synthesis underpins evidence-based medicine, public health, environmental science, and policy - but many widely used tools are proprietary, existing directories mix software with checklists and guidance documents, licensing is often unclear, and limited transparency restricts reproducibility. This project fills those gaps with a strictly open-source-only, software-focused directory with verified repositories and licenses - a discovery platform for reusable research software supporting Open Science and FAIR principles, and a foundation for interoperable, transparent tools.
Only tools meeting all criteria are included:
- Open source license - recognized license (MIT, GPL, Apache, etc.)
- Public code repository - GitHub, GitLab, Bitbucket, or SourceForge
- Non-proprietary - no closed-source components, mandatory commercial dependencies, or institutional license requirements
- Reusable & extensible - documentation must allow reuse, modification, and community development
- Research-relevant - related to evidence synthesis, systematic reviews, meta-analysis, or closely related workflows (including general-purpose libraries used for ES automation)
Excluded: EndNote, Covidence, DistillerSR, NVivo, Rayyan, Stata, SAS, and similar proprietary/freemium tools; also tools whose code is hosted exclusively on university/institutional websites (accessibility and long-term stability concerns).
External API policy: open-source tools that interact with external databases, platforms, or APIs (e.g., literature databases or screening systems) are eligible if the software itself is fully open source with a public repository, its core functionality is transparent and reusable, and the external service is used only for data access or integration - not hidden proprietary logic. These are documented as open-source integrations with external research infrastructure.
- Transparency - inspect underlying algorithms
- Reproducibility - avoid black-box systems
- Sustainability - independence from commercial vendors
- Innovation - fork, modify, and build 🚀
- Equity - global access without paywalls
Unlike existing directories such as the Systematic Review Toolbox (SR Toolbox) - which includes proprietary and mixed content, has unclear source code availability and licensing, limited update transparency, and no public source for its own platform - this project applies a strict open-source-only policy with verified repositories and licenses, a software-only focus, and pure HTML for easy maintenance. The author initially tried to contribute new tools to existing platforms, but their closed source prevented transparency, extensibility, and community contribution. The result enables developers to inspect, extend, and build upon real-world evidence synthesis implementations - discovering fully open codebases and reusing components like NLP libraries, PDF parsers, and ML frameworks - aligned with Open Science and research software sustainability principles.
The website is intentionally plain HTML: no framework or dependency lock-in, long-term stability and preservation, easy editing in any text editor, and contributions possible with little or no coding experience.
Submit a tool via Pull Request (preferred), Issue, Discussions, or email the maintainer (V.S.).
Adding a tool (counts update automatically): all tool numbers - section counts, the directory-wide total, and the header badge - are computed from the tool cards in index.html on page load. Copy an existing .tool-card block into the appropriate section (#section-core, #section-methodologists, or #section-developers); no manual counting required.
Requirements: open-source license, public repository, no closed-source components, no paid or institutional restrictions, relevance to evidence synthesis, adequate documentation. Not accepted: free but closed-source tools, freemium software, commercial or license-restricted platforms.
This project was conducted independently by the author. AI tools supported interface refinement and presentation only - GLM-4.5V-Flash (open access) improved layout structure and visual clarity. All conceptual development, inclusion criteria, repository/license verification, tool selection, curation, and project design were performed manually by the author. More info: https://doi.org/10.5281/zenodo.18692063, evidencesynthesis-tools/backwork, https://doi.org/10.31222/osf.io/7uskw_v1.
The initial systematic search was conducted in October 2025, with continued searches and verification covering tools available up to February 2026. The directory is complete up to early 2026, as defined by the associated mapping study; all searches and verification were performed manually by the author. Maintenance beyond 2026 is at the author's discretion - the directory will be expanded only if there is a genuine, demonstrated need (sustained community use, active contributors, or clear demand). Contributors and collaborators are welcome to support long-term maintenance.
Sahu, V. (2026). Evidence Synthesis Tools: A curated directory of strictly open-source software / tools for Evidence Synthesis. (Version 5.0). Zenodo. https://doi.org/10.5281/zenodo.18681868
Individual tools retain their original citations.
© 2026 Vihaan Sahu - Apache 2.0 License.
Built for the research community to advance open, transparent, and reproducible evidence synthesis 🌐