DEV.co
MCP Servers · papersgpt

papersgpt-for-zotero

PapersGPT is a Zotero plugin that integrates multiple AI models (ChatGPT, Claude, Gemini, DeepSeek, and others) to help researchers chat with, summarize, and analyze PDF documents directly within their Zotero library. It supports both cloud-based LLMs and local open-source models, with batch processing capabilities through an AutoPilot feature.

Source: GitHub — github.com/papersgpt/papersgpt-for-zotero
2.5k
GitHub stars
85
Forks
JavaScript
Primary language
AGPL-3.0
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositorypapersgpt/papersgpt-for-zotero
Ownerpapersgpt
Primary languageJavaScript
LicenseAGPL-3.0 — OSI-approved
Stars2.5k
Forks85
Open issues73
Latest releasepapersgpt-v0.6.0 (2026-07-03)
Last updated2026-07-03
Sourcehttps://github.com/papersgpt/papersgpt-for-zotero

What papersgpt-for-zotero is

JavaScript-based Zotero extension providing RAG-enabled document analysis via multiple LLM backends (OpenAI, Anthropic, Google, custom APIs, OpenRouter, and local inference via Ollama/local runners). Includes local embeddings, vector storage, and re-ranking; MCP protocol support for external chatbot integration; batch automation via AutoPilot dashboard.

Quickstart

Get the papersgpt-for-zotero source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/papersgpt/papersgpt-for-zotero.gitcd papersgpt-for-zotero# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Literature Review at Scale

Process 100+ papers overnight with AutoPilot, extracting methodology, findings, and custom data points. Results sync directly to Zotero notes, eliminating manual copy-paste workflows.

Privacy-First Research Analysis

Researchers handling sensitive or proprietary documents can run entirely local inference (Gemma, Qwen, DeepSeek distilled models) with no data transmission, suitable for compliance-heavy environments.

Multi-Model Experimentation

Academic teams and AI researchers can quickly compare outputs across different LLMs (via OpenRouter or direct APIs) without switching tools, using a single plugin interface.

Implementation considerations

  • API key management: Users must configure and secure API keys for chosen LLM provider(s) within the plugin; no centralized key store documented.
  • Local GPU/CPU: Ensure adequate hardware if running local models; plugin auto-selects GPU when available, but resource contention with Zotero may occur on constrained machines.
  • Network stability: Cloud LLM usage and local model downloads depend on robust internet; Hugging Face and GitHub access required for local model fetching.
  • Batch processing limits: AutoPilot has undocumented daily caps during beta; validate expected throughput against organization's processing volume.
  • Zotero library state: Plugin modifies Zotero notes directly; ensure backups and test on non-production libraries first.

When to avoid it — and what to weigh

  • Zotero Versions Below 8 — Plugin requires Zotero 8 or 9 (v0.3.7+). Older Zotero deployments will not be supported.
  • Strict Data Residency / Offline-Only Workflows — While local LLMs are supported, cloud model options (OpenAI, Anthropic, Gemini) require internet connectivity and sending document content to external APIs. Organizations with absolute no-cloud policies should evaluate local-only setup carefully.
  • Non-English or Heavily Formatted Documents — No documented support for OCR, scanned images, or non-Latin scripts. Plugin assumes digital, text-extractable PDFs.
  • Production SLA Requirements — AutoPilot is in 'early testing phase with daily usage limits.' Not suitable for time-critical, production research pipelines requiring guaranteed uptime.

License & commercial use

Licensed under AGPL-3.0 (GNU Affero General Public License v3.0). This is a strong copyleft license requiring that any derivative works or network-accessible modifications must also be open-sourced under AGPL-3.0. Modifications and improvements must be distributed under the same license.

AGPL-3.0 permits commercial use, but imposes significant obligations: any modifications or deployment as a service (SaaS) must provide source code to users and comply with copyleft requirements. Commercial vendors should conduct legal review before integrating or modifying this plugin. Using the unmodified plugin for internal research is lower-risk than packaging or reselling it.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationAdequate
License clarityClear
Deployment complexityLow
DEV.co fitGood
Assessment confidenceHigh
Security considerations

API keys are user-configured and stored in plugin settings; no documented encryption or key rotation guidance. Local models and RAG (embeddings, vector DB) run on-machine, limiting data exfiltration risk for local-only setups. Cloud model usage sends document content to external providers (OpenAI, Anthropic, Google, etc.); users must review those vendors' security/privacy policies. No published security audit, vulnerability disclosure policy, or incident response process documented. Requires Zotero trust and JavaScript execution privileges.

Alternatives to consider

Zotero built-in note features + ChatGPT/Claude web UI

No plugin required; manually copy PDFs to ChatGPT or Claude web interface. Simpler for low-volume use, no local setup, but loses Zotero integration and batch automation.

Semantic Scholar / Elicit (AI research tools)

Purpose-built for literature discovery and summarization; integrated with academic databases. No Zotero sync, but may offer better metadata parsing and domain-specific models.

Obsidian with PDF plug-ins + LLM integrations

Flexible note-taking with PDF support and AI add-ons. Requires manual PDF import; less integrated with research libraries than Zotero, but more open architecture.

Software development agency

Build on papersgpt-for-zotero with DEV.co software developers

Download PapersGPT v0.6.0 and connect your preferred AI model to Zotero today. Start with local privacy-first LLMs or cloud APIs—your choice.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

papersgpt-for-zotero FAQ

Does PapersGPT store my documents or send them to PapersGPT servers?
When using cloud LLMs (ChatGPT, Claude, Gemini), documents are sent to those providers' APIs. Local LLM mode keeps data on-machine. No central PapersGPT server storage is documented, but users should verify with their chosen LLM provider's privacy policy.
What are the AutoPilot daily usage limits during beta?
Limits are not specified in available documentation. Contact the project or check GitHub issues for current beta rate limits and expected GA timeline.
Can I use PapersGPT offline?
Yes, if running local LLMs (Gemma, Qwen, DeepSeek distilled). Cloud LLM use requires internet. Model downloads and initial setup require network access.
Is there a way to batch-process papers without AutoPilot?
AutoPilot is the documented batch feature. Manual multi-PDF chat is supported but not batch-scripted. Script automation via SKILL integration may be possible for advanced users.

Software development & web development with DEV.co

Adopting papersgpt-for-zotero is usually one piece of a larger software development effort. As a software development agency, DEV.co provides software development services and web development expertise — pairing senior software developers and web developers with your team to design, build, and operate mcp servers software in production.

Ready to Accelerate Your Literature Review?

Download PapersGPT v0.6.0 and connect your preferred AI model to Zotero today. Start with local privacy-first LLMs or cloud APIs—your choice.