| name | arxiv-reader |
| description | Read, summarize, or explain any arxiv paper. Supports arxiv.org URLs (abs, pdf, html), paper IDs, and alphaxiv.org URLs. Use when user shares an arxiv link, mentions a paper ID, or asks about a research paper. |
Arxiv Reader
Read, summarize, or explain any arxiv paper. Default approach fetches the arxiv HTML page directly via WebFetch, with alphaxiv.org as a supplementary source for structured overviews.
When to Use
- User shares any arxiv URL (e.g.
arxiv.org/abs/2401.12345, arxiv.org/pdf/2401.12345, arxiv.org/html/2603.02473v1)
- User mentions a paper ID (e.g.
2401.12345, 2603.02473v1)
- User asks you to read, explain, summarize, or analyze an arxiv research paper
- User shares an alphaxiv URL (e.g.
alphaxiv.org/overview/2401.12345)
- User pastes a link to a paper and wants to understand it
Workflow
Step 1: Extract the paper ID
Parse the paper ID from whatever the user provides:
| Input | Paper ID |
|---|
https://arxiv.org/abs/2401.12345 | 2401.12345 |
https://arxiv.org/pdf/2401.12345 | 2401.12345 |
https://arxiv.org/html/2603.02473v1 | 2603.02473v1 |
https://alphaxiv.org/overview/2401.12345 | 2401.12345 |
2401.12345v2 | 2401.12345v2 |
2401.12345 | 2401.12345 |
Step 2: Fetch the paper via arxiv HTML (primary)
Use WebFetch to get the arxiv HTML page directly. This works for all papers including newly published ones:
WebFetch(url="https://arxiv.org/html/{PAPER_ID}", selector="article", no_js=true)
selector="article" extracts the main paper content, skipping navigation and boilerplate.no_js=true since arxiv HTML pages are static — no JavaScript rendering needed.- This is the default and preferred method. It always has the latest papers.
If the HTML version is not available (some older papers only have PDF), proceed to Step 3.
Step 3: Fetch alphaxiv overview (supplementary)
alphaxiv.org provides AI-generated structured overviews, but may not have new papers yet.
curl -s "https://alphaxiv.org/overview/{PAPER_ID}.md"
Returns a structured, detailed analysis optimized for LLM consumption. One call, plain markdown, no JSON parsing.
If this also returns 404, try the full text extraction:
curl -s "https://alphaxiv.org/abs/{PAPER_ID}.md"
Step 4: Last resort — PDF
If both HTML and alphaxiv are unavailable, direct the user to the PDF:
https://arxiv.org/pdf/{PAPER_ID}
Error Handling
- arxiv HTML unavailable: Some older papers lack HTML versions. Fall through to alphaxiv.
- alphaxiv 404: Paper not yet processed by alphaxiv. Expected for very new papers.
- Both unavailable: Point user to the PDF URL.
Notes
- No authentication required — all endpoints are public.
- Prefer arxiv HTML over alphaxiv because arxiv has zero latency for new papers.
- Use alphaxiv when you want its AI-generated structured overview as a supplement, not as the primary source.
