README file from
GithubResearch Paper
AI-assisted research briefs from academic databases — right inside Obsidian.
Quick start
- Install from Community Plugins → search "Research Paper"
- Enable it in Settings → Community Plugins
- Open Settings → Research Paper, pick an LLM provider, paste your API key
- Open a note, click the 🔍 ribbon icon, and start researching
The plugin auto-detects your Obsidian language.
LLM Providers
| Provider | Models |
|---|---|
| OpenAI | GPT-5.5, GPT-5.4, GPT-5.4 Mini, GPT-5.4 Nano, GPT-5.2 |
| Anthropic | Claude Opus 4.7, Sonnet 4.6, Haiku 4.5 |
| DeepSeek | V4 Pro, V4 Flash |
| Google Gemini | 2.5 Pro, 2.5 Flash, 2.5 Flash Lite |
| OpenRouter | Routes to all major models |
| Grok (xAI) | Grok 4.3, Grok 4.20 Reasoning |
| GLM (Z.ai) | GLM-4 Plus, GLM-4, GLM-4 Flash, GLM-4 Air |
Tutorial
1. Configure your LLM provider
Go to Settings → Research Paper. Select your provider (e.g., OpenAI), paste your API key, and pick a model. GPT-5.5 and Claude Opus 4.7 produce the best results.
Optionally, add a PubMed API Key (free, increases search rate limits) and your email for Crossref DOI verification.
2. Search for evidence
Open a note, click the 🔍 ribbon icon or run Ctrl+P → Research Paper: buscar y generar. Type your research question in natural language — the LLM optimizes it into academic keywords automatically:
"What's the evidence on CBT for generalized anxiety disorder?"
"¿Qué dice la literatura sobre terapia cognitivo-conductual en adolescentes con depresión?"
Choose your domain (10 options, from Psychology to Engineering), mode (Quick Answer or Research Brief), language for the output, and a year range filter.
3. Review and refine
After searching, you'll see a list of articles with relevance scores. Each one is pre-selected with a checkbox:
- Uncheck articles that aren't relevant to your focus
- Use the Extra instructions field to guide the AI in natural language:
- "Focus on efficacy compared to pharmacotherapy"
- "Only include randomized controlled trials"
4. Generate your brief
Click Generate brief. The output is a structured Research Brief inserted into your note:
## Research Brief: {Title}
⚠️ AI-assisted brief. All factual data is traceable to the listed sources.
### Research Question
{Your original query}
### Sources Consulted
1. Smith, J. et al. (2025). *Journal*. DOI: [10.xxx](https://doi.org/10.xxx)
{Summary in your language}
### Findings per Source
→ Smith et al. (2025) found that...
### Convergences
• Both studies agree that... (Smith, 2025; García, 2024)
### Divergences
• Smith reports a large effect; García reports moderate...
### Implications
{AI-generated synthesis}
⚠️ This section is AI-generated synthesis.
Features
- Natural language search — optimized into academic keywords by LLM
- Three search sources — PubMed (biomedical domains), OpenAlex, and Semantic Scholar (214M+ papers)
- Research Brief format — traceable evidence, not fake papers. No fabricated Methods, no false claims
- Review & Refine — select exactly which articles to include, add custom instructions
- APA 7 in-text citations — consistent author-year format throughout
- DOI verification — validates against Crossref API
- Year range filter — All time / Last 3 / 5 / 10 / 20 years
- 10 research domains — Psychology, Medicine, Nursing, Biology, Education, Sociology, Economics, Law, Engineering, General
- 6 output languages — with mandatory English abstract
- Model selector — 30+ models across 7 providers, updated for 2026
Search sources
| Database | Coverage | Requires |
|---|---|---|
| PubMed | Biomedical & clinical | None (API key optional) |
| OpenAlex | Multi-disciplinary, 250M+ works | None |
| Semantic Scholar | Multi-disciplinary, 214M+ papers | None |
Support
Research Paper is free and open source. If it helps your academic work, consider supporting development:
Credits
Created by Jesús García & DeepSeek V4-Pro · GitHub