ReviewAid

ReviewAid is a free, open-source tool for AI-assisted full-text screening and extraction in systematic reviews. It is best for researchers willing to configure their review and check the results, rather than delegate decisions to software. Its strongest case is configurable assistance with a local-processing option.

Overview

ReviewAid applies user-defined PICO criteria to full-text papers and lets researchers specify which fields to extract. Confidence scores and low-confidence flags can help focus manual checks, but the maker explicitly positions the tool as supplementary to human screening and extraction. Reference deduplication is supported; blinded screening is not, which may rule it out for workflows that require blinding.

The Apache 2.0 license permits use as open-source software. Researchers can use a browser-based Streamlit app without installation or run it locally. The hosted app may take about 30 seconds to initialize, so it is less suited to users who expect immediate access on every visit.

Key features

Screening and extraction

Supplying PICO inclusion and exclusion criteria gives reviewers a way to tailor full-text screening to a review question, while custom extraction fields accommodate different evidence-synthesis needs. That flexibility comes with setup and checking work: the tool is an assistant, not an authority on inclusion decisions or extracted data.

Confidence, exports, and batch limits

ReviewAid flags low-confidence results and offers CSV, XLSX, and DOCX exports, making it practical to review uncertain decisions and carry results into other tools. Submission sizes are modest: the README advises up to 10 articles per full-text screening submission and up to 20 for extraction. The site's batch-processing description also refers to up to 20 papers, so screening users should follow the lower limit.

Provider and privacy choices

Supported providers include OpenAI, Anthropic, DeepSeek, Cohere, Z.ai, and local Ollama. With a cloud provider, paper text is sent to that provider for the run; the maker says API keys are not stored. Ollama runs inference on the local machine without an external API key, keeping paper text there. That is the clearest fit for researchers who prioritize local processing, though it requires running ReviewAid locally.

Pricing

Free: free. The plan is open source under Apache 2.0, with no paid tier described. There is no subscription price to weigh against the limited batch sizes, but users should account for their own setup and the work of checking model results. Cloud-provider use sends paper text to the selected provider; the local Ollama route avoids that transfer.

Platforms

ReviewAid supports Linux, macOS, Windows, web, and self-hosted use. The browser-based Streamlit app avoids installation; local installation uses Python 3.12 or later and Streamlit on Windows, macOS, or Linux. Choose the hosted route for convenience, bearing in mind possible cold starts, or local deployment for more control over processing.

Who it's for

ReviewAid suits researchers conducting systematic reviews, particularly evidence-synthesis teams that want configurable AI assistance and can maintain human oversight. It is a poor fit for projects that require blinded screening or for reviewers who want to hand off screening and extraction without checking results.

Pros and cons

Pros: PICO-based screening and custom extraction fields can be adapted to a review's question and data needs.

Pros: Confidence flags and CSV, XLSX, and DOCX exports support targeted checking and downstream use.

Pros: Local Ollama inference keeps paper text on the machine and does not require an external API key.

Cons: Screening submissions are limited to 10 articles, which constrains work in larger batches.

Cons: Blinded screening is absent, excluding workflows that depend on it.

Cons: The hosted app can take about 30 seconds to start, while local use requires Python 3.12 or later and Streamlit.

Alternatives

For a broader comparison, see Systematic Review Software.

Universal SLR Assistant is another free, open-source option for researchers comparing tools in this category.

ASReview is also free and open source, with self-hosted and API platforms; consider it if you want to compare another no-subscription option.

Rayyan has a free plan capped at three active reviews and two free reviewers; choose it if those limits suit your team and you want its Android or iOS access.

Parsifal offers free web access and an optional AI assistant using your own Anthropic or OpenAI API key; consider it if that arrangement fits your workflow.

Silvi's free Basic plan promises unlimited reviews, easy screening, and transparent extraction; consider it if unlimited review count matters more than ReviewAid's local-inference option.

Bango is free for researchers with no subscription, paywall, or limits; compare it if avoiding usage limits is your priority.

RevMan is a freemium option with a free plan and trial.

Buhos is another free option for comparison.

Verdict

Choose ReviewAid if you conduct systematic reviews, want configurable full-text screening and extraction at no subscription cost, and will verify the results. Its strongest differentiator is the option to keep paper text local with Ollama. Look elsewhere if blinded screening is essential, larger screening batches are needed, or you expect the software to replace human review.