ASReview Free plan 9.1 Visit Universal SLR Assistant
sadeghanisi.github.io
What it is
Universal SLR Assistant supports systematic and scoping reviews with AI-assisted screening, extraction, and reproducible outputs. It provides a desktop interface and a browser-based local Web App, and imports RIS, BibTeX, CSV, and plain-text references, including exports from PubMed, Scopus, and Web of Science. Its workflow covers abstract screening, full-text PDF processing, structured extraction, and export; likely duplicates are identified by DOI and fuzzy title matching. Workspace mode stores references, PDF metadata, review items, decisions, and audit events in a local SQLite workspace, separating AI suggestions from human decisions. Exports can include decision files and logs, exclusion reasons, PRISMA-ready counts, a methods disclosure draft, and a manifest. The project supports a range of cloud and local AI providers. Cloud providers receive submitted paper text, while Ollama or another local endpoint can keep it on the user's computer or local network. AI results are assistance only: people must verify them and make final review and reporting decisions. The software is free under the MIT License; cloud AI provider API fees may apply.
Spec sheet
- What it does
- Universal SLR Assistant helps researchers import references, deduplicate records, screen literature, process PDFs, extract structured data and export auditable review outputs.Read 15 Oct 2026 · sadeghanisi.github.io
- Purpose
- Universal SLR Assistant supports systematic and scoping review workflows with AI-assisted screening, extraction, and reproducible outputs.Read 30 Sep 2026 · github.com
- Intended users
- It is intended for PhD students, researchers, professors, evidence synthesis teams, and systematic or scoping review authors.Read 30 Sep 2026 · sadeghanisi.github.io
- Privacy
- The Web App binds to 127.0.0.1 for local use and is not intended as a public multi-user service.Read 30 Sep 2026 · github.com
- Support
- The project directs users to the Complete User Guide, in-app Help panel, output logs, tests, and GitHub issues for support.Read 30 Sep 2026 · github.com
- Abstract screening
- AI evaluates titles and abstracts against user criteria and returns Include, Exclude or Maybe recommendations for human review.Read 15 Oct 2026 · sadeghanisi.github.io
- AI providers
- Provider options include OpenAI, Anthropic Claude, Google Gemini, Ollama, OpenRouter, DeepSeek, Mistral, Kimi, xAI, LM Studio, vLLM, LocalAI, and custom OpenAI-compatible endpoints.Read 30 Sep 2026 · github.com
- API and extensibility
- The Web App exposes workspace review and export HTTP endpoints, and contributors can add providers through the LLM interface or OpenAI-compatible profiles.Read 15 Oct 2026 · github.com
- Audit ledger
- Each run can create a JSONL ledger containing provider, model, hashes, cache status, parsing status and retry metadata without storing API keys or full paper text.Read 15 Oct 2026 · sadeghanisi.github.io
- Caching
- Configuration-aware JSON caching keys results by text, stage, provider, model, prompts and extraction settings.Read 15 Oct 2026 · github.com
- Cloud data transmission
- When cloud providers are selected, submitted paper text is sent to the provider's API; router providers add another intermediary.Read 15 Oct 2026 · sadeghanisi.github.io
- Data transmission
- Cloud providers receive submitted paper text, while Ollama or another local endpoint can keep paper text on the user's computer or local network.Read 30 Sep 2026 · sadeghanisi.github.io
- Deduplication
- Duplicate detection combines DOI matching with fuzzy title matching and leaves final review to the researcher.Read 15 Oct 2026 · sadeghanisi.github.io
- Developer post: automation
- A mirrored developer post lists automated deduplication and extraction as product capabilities.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer post: cost
- A mirrored developer post says the tool has no monthly fees and is free under the MIT license.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer post: local integration
- A mirrored developer post highlights local Ollama integration and processing PDFs on the user's own machine.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer post: platforms
- A mirrored developer post names Windows, macOS and Linux as supported desktop platforms.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer post: privacy
- A mirrored developer post claims local LLM use can keep data from leaving the user's hardware.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer post: providers
- A mirrored developer post describes support for more than nine AI providers, including OpenAI, Claude, Gemini and DeepSeek.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Developer profile
- A public profile identifies Mo Anisi as holding a BSc in Biology, an MSc in Cognitive Science and an MSc in Health and Social Welfare.Read 15 Oct 2026 · ngntipkolamrenang.twstalker.com
- Documentation
- Documentation includes a 1,550-line complete user guide, in-app help, troubleshooting guidance, tests and GitHub issues.Read 15 Oct 2026 · github.com
- Exports
- Workspace exports include decision CSV/XLSX files, AI and human decision logs, exclusion reasons, PRISMA-ready counts, a methods draft and a manifest.Read 15 Oct 2026 · github.com
- Hardware guidance
- The user guide lists 4 GB RAM and 2 GB disk as minimums, with 8 GB RAM and 5 GB disk recommended; Ollama guidance lists at least 8 GB RAM and 10 GB disk.Read 15 Oct 2026 · github.com
- Human decisions
- AI outputs remain suggestions, while final Include, Exclude or Maybe decisions require a human reviewer.Read 15 Oct 2026 · github.com
- Installation
- Installation uses a Git clone, virtual environment and pip requirements, with batch launchers for Windows and Python commands for macOS or Linux.Read 15 Oct 2026 · sadeghanisi.github.io
- Interfaces
- The project provides a desktop GUI and a browser-based local Web App.Read 30 Sep 2026 · sadeghanisi.github.io
- Limit: appraisal
- The product does not provide risk-of-bias or quality-appraisal tooling.Read 15 Oct 2026 · github.com
- Limit: collaboration
- The changelog says Workspace Mode is single-reviewer only and has no dual-reviewer, conflict-adjudication or kappa workflow.Read 15 Oct 2026 · github.com
- Limit: extraction persistence
- Structured extraction remains run-based and is not yet persisted or reviewable in the workspace database.Read 15 Oct 2026 · github.com
- Limit: OCR
- The changelog states that scanned-PDF OCR is not available.Read 15 Oct 2026 · github.com
- Limit: validation
- The project documents that it has no formal real-world LLM validation benchmark yet.Read 15 Oct 2026 · github.com
- Limitations
- AI output is assistance only; human verification and final inclusion, exclusion, extraction, and reporting decisions remain required.Read 30 Sep 2026 · sadeghanisi.github.io
- Local inference
- Ollama and other local endpoints can keep paper text on the user's computer or local network, depending on hosting.Read 15 Oct 2026 · sadeghanisi.github.io
- PDF processing
- The tool uses a multi-engine PDF extraction cascade and records extraction status.Read 15 Oct 2026 · sadeghanisi.github.io
- Provider integrations
- Provider options include OpenAI, Anthropic Claude, Google Gemini, Ollama, OpenRouter, DeepSeek, Mistral, Kimi, xAI, LM Studio, vLLM, LocalAI and custom OpenAI-compatible endpoints.Read 15 Oct 2026 · sadeghanisi.github.io
- Reference imports
- It accepts RIS, BibTeX and CSV references, including exports from PubMed, Scopus and Web of Science.Read 15 Oct 2026 · sadeghanisi.github.io
- Reference ingestion
- It imports RIS, BibTeX, CSV, and plain-text reference files and supports exports from PubMed, Scopus, and Web of Science.Read 30 Sep 2026 · github.com
- Release news
- GitHub lists version 3.5.3 as released on 11 October 2026 with reliability and workspace data-integrity updates.Read 15 Oct 2026 · github.com
- Review types
- The software supports systematic and scoping review workflows.Read 15 Oct 2026 · github.com
- Review workflow
- The workflow covers reference import, abstract screening, full-text PDF processing, structured extraction, and export.Read 30 Sep 2026 · sadeghanisi.github.io
- Security control
- The desktop and Web App settings files do not save API keys, and workspace metadata scrubs secrets and full paper text.Read 15 Oct 2026 · github.com
- Security controls
- API keys are not written to the desktop settings.json or Web App webapp_settings.json, and audit ledgers do not store API keys or full paper text.Read 30 Sep 2026 · github.com
- Structured extraction
- Users define extraction fields and receive Excel or CSV drafts for checking and reporting.Read 15 Oct 2026 · sadeghanisi.github.io
- System requirements
- The current website lists Python 3.10 or newer, Flask, a modern browser and either an AI-provider key or Ollama/local endpoint.Read 15 Oct 2026 · sadeghanisi.github.io
- Two-stage screening
- A fast first screening stage can be followed by detailed full-text criteria screening.Read 15 Oct 2026 · sadeghanisi.github.io
- Workspace mode
- Workspace mode stores references, PDF metadata, review items, decisions, and audit events in a local SQLite workspace and separates AI suggestions from human decisions.Read 30 Sep 2026 · github.com
- Workspace storage
- Workspace Mode stores references, PDF metadata, review items, decisions and audit events in a local SQLite database.Read 15 Oct 2026 · github.com
Plans and prices
Free and open source Free MIT License · cloud AI provider API fees may apply Read 15 Oct 2026 · sadeghanisi.github.io
Strong points and limits
Strong points
- Imports RIS, BibTeX, CSV, and plain-text references
- Flags likely duplicates using DOI and fuzzy title matching
- Stores workspace data in local SQLite
- Exports PRISMA-ready counts and decision logs
Limits
- Cloud providers receive submitted paper text
- Cloud AI provider API fees may apply
- AI output requires human verification and decisions
Who it suits
It is intended for PhD students, researchers, professors, evidence synthesis teams, and authors of systematic or scoping reviews. It suits users who want structured review workflows while retaining human control over final decisions.
Universal SLR Assistant in depth
Read the full write-up
Universal SLR Assistant is a free tool for conducting systematic and scoping reviews, aimed at researchers who need screening, extraction and reporting in one workflow. Its best fit is someone who values a local, auditable workspace and a choice of AI providers. It is less suitable for teams seeking a public, multi-user web service or automated final decisions.
Overview
The workflow spans reference import, abstract screening, full-text PDF processing, structured extraction and export. It accepts RIS, BibTeX, CSV and plain-text files, including exports from PubMed, Scopus and Web of Science. DOI and fuzzy-title matching flag likely duplicates before screening, helping reduce repeated records while leaving users responsible for checking matches.
Workspace mode keeps references, PDF metadata, review items, decisions and audit events in a local SQLite workspace. Crucially, AI suggestions remain separate from human decisions, a useful distinction for evidence synthesis where recommendations must not be mistaken for adjudication. Exports include decision files, separate AI and human decision logs, full-text exclusion reasons, PRISMA-ready counts, a methods disclosure draft and a manifest. That combination favors structured, reproducible reviews over casual paper discovery.
Key features
AI assistance across the review
Provider options include OpenAI, Anthropic Claude, Google Gemini, Ollama, OpenRouter, DeepSeek, Mistral, Kimi, xAI, LM Studio, vLLM, LocalAI and custom OpenAI-compatible endpoints. This breadth gives users flexibility to work with cloud providers or local endpoints. It does not remove the need to verify screening and extraction: people retain responsibility for inclusion, exclusion, extraction and reporting decisions.
Local workspace and privacy choices
The browser-based Web App binds to 127.0.0.1 for local use and is not intended to serve multiple public users. API keys are not written to the desktop or Web App settings files, and audit ledgers omit both keys and full paper text. Cloud providers receive paper text submitted to them; using Ollama or another local endpoint can keep that text on the user's computer or local network. The distinction matters for sensitive material, though local processing depends on choosing and configuring a local endpoint.
Pricing
The Free and open source plan costs nothing under the MIT License. It includes screening, AI paper summaries, full-text analysis, systematic review support, deduplication, data extraction and full-text storage, with exports in Excel (.xlsx), CSV, JSON, Markdown and plain text. No software fee is charged, but cloud AI provider API fees may apply, so free access to the application does not necessarily mean cost-free use.
This plan suits individual researchers and evidence synthesis teams comfortable managing their own workflow and any provider charges. Support routes include the Complete User Guide, in-app Help panel, output logs, tests and GitHub issues. AI remains assistance only, so users who need final decisions made without human verification should choose another approach.
Platforms
Universal SLR Assistant supports Linux, macOS, Windows and the web. The project offers both a desktop GUI and a browser-based local Web App; the latter's localhost binding makes it a personal local interface rather than a public collaboration service.
Who it's for
PhD students, researchers, professors, evidence synthesis teams and authors of systematic or scoping reviews are the intended audience. It is particularly well matched to users who need a trail of decisions and reporting-oriented exports, and who are willing to verify model suggestions. It is a weaker fit for groups that require a publicly hosted shared workspace.
Pros and cons
Pros
End-to-end review workflow: import, screening, full-text processing, extraction and export are covered in one tool.
Auditable work: local workspace records review activity and keeps AI suggestions distinct from human decisions.
Useful reporting outputs: decision logs, exclusion reasons, PRISMA-ready counts and a methods disclosure draft support reproducible reporting.
Provider choice: cloud and local endpoint options let users choose how paper text is processed.
No software charge: the MIT-licensed plan costs nothing, although cloud API use can add fees.
Cons
Human verification remains essential: the software does not make final inclusion, exclusion, extraction or reporting decisions.
Cloud use transmits paper text: users handling sensitive material need to choose a local endpoint if they want text to stay on their computer or network.
Web App is not a shared service: its local-only binding rules it out as a public multi-user workspace.
Potential provider costs: cloud AI fees sit outside the free software plan.
Alternatives
Rayyan is worth considering for teams that prefer a freemium service with a free tier capped at three active reviews and two free reviewers; its listed platforms include Android, iOS, web and API.
ASReview is a free, open-source alternative for users who favor running review work on their own hardware or server, which may incur costs; it supports Linux, macOS, Windows, web, API and self-hosting.
Moara may suit users who want a web-based review library with Google Scholar and Semantic Scholar search and collaboration for viewers or editors; its free plan allows up to 50 papers per library.
RobotReviewer is another free, open-source option.
Scholarcy is a more focused choice for article summaries: its free plan allows 10 summaries and one-at-a-time flashcard exports.
Elicit may fit researchers looking for broad paper search, summaries and chat, with limited use of its Research Agent and Research Reports on the Basic plan.
Axelium offers a free plan limited to one lifetime review and one seat, which may suit someone who wants to try its core features on a single project.
Silvi is an alternative for users seeking a web tool whose free Basic plan offers unlimited reviews, screening and data extraction.
For broader comparisons, see AI Literature Review Software and Systematic Review Software.
Verdict
Choose Universal SLR Assistant if you want a free, full-cycle systematic or scoping review workflow with a local workspace, decision trail and flexible AI-provider options. Its strongest advantage is the combination of reproducibility-oriented exports and separation of model suggestions from human decisions. Look elsewhere if you need a public collaborative web service, or expect AI to make final review judgments for you.
Verdict
Universal SLR Assistant offers a broad review workflow, local workspace storage, and several export options at no software cost. Users should account for possible provider fees and retain responsibility for verifying outputs and making final decisions.
Where Specifiction places it
- Systematic Review Software#6 of 308.2
- AI Literature Review Software#10 of 388.2
