Show HN: RenderarXiv – Search ArXiv from terminal, HTML to read/paste into LLM

2 hours ago 1

Search arXiv from your terminal and get beautiful HTML results you can read or copy-paste into ChatGPT/Claude.

No more hallucinated papers — everything is real, straight from arXiv.

pip install git+https://github.com/peterdunson/renderarxiv.git

That's it!

renderarxiv "transformer attention mechanism"

Opens a beautiful HTML page in your browser with:

👤 Human view — clean paper cards with abstracts, authors, PDF links
🤖 LLM view — formatted text ready to copy into AI assistants

Get more/fewer results:

renderarxiv "quantum computing" --max-results 15

Ranking modes:

renderarxiv "deep learning" --mode recent # Newest papers renderarxiv "neural networks" --mode relevant # Best text match renderarxiv "language models" --mode semantic # Smart semantic matching

Default is balanced (mix of relevance + recency).

Filter by category:

renderarxiv "object detection" --category cs.CV # Computer Vision only renderarxiv "optimization" --category math.OC # Math only

Common categories: cs.LG (ML), cs.AI (AI), cs.CL (NLP), cs.CV (Vision), cs.RO (Robotics)

Save to file:

renderarxiv "diffusion models" -o papers.html --no-open

# Latest ML research renderarxiv "large language models" --category cs.LG --mode recent # Find a specific paper renderarxiv "attention is all you need" --mode relevant # Explore robotics renderarxiv "robot manipulation" --category cs.RO --max-results 20 # Deep semantic search renderarxiv "few-shot learning" --mode semantic

Search for papers: renderarxiv "your topic" --max-results 10
Click the 🤖 LLM button in your browser
Copy the text (Ctrl+A, Ctrl+C)
Paste into ChatGPT/Claude: "Summarize these papers and identify key trends"

Now your AI has real citations, not hallucinations!

Computer Science: