arxiv — The Research Fang

Knowledge is the only weapon that sharpens itself. This B1TE cuts through the noise.

Part of FRE4X-B1TE — a monorepo of MCP servers built for autonomous agents.

Agents sleep while humans read. This server doesn't sleep. It hunts across the entire arXiv corpus — millions of papers, every category, in real time — so the agent always knows what humans don't yet.

Tools

| Tool | What it hunts | |------|--------------| | arxiv_search_papers | Free-text search with field prefixes (ti:, au:, abs:, cat:) and boolean operators | | arxiv_get_paper | Full metadata for one or more papers by arXiv ID | | arxiv_search_by_author | All papers by a specific author | | arxiv_search_by_category | Papers in a subject category (e.g. cs.AI, cs.LG, q-fin.TR) | | arxiv_list_categories | All supported arXiv subject categories |

Query Syntax

ti:attention AND cat:cs.LG
au:Bengio AND abs:generalization
cat:cs.CV AND ti:object detection
ti:diffusion AND NOT abs:image
all:"Scaling Law limits" AND submittedDate:[20230101 TO 20261231]

All list tools support limit (max 100) and offset for pagination.
All tools support response_format: "markdown" (default) or "json".

Response Examples

arxiv_search_papers

Input

{
  "query": "all:\"Scaling Law limits\" AND submittedDate:[20230101 TO 20261231]",
  "sort_by": "relevance",
  "limit": 5
}

Output (markdown)

# arXiv Search Results

Showing 5 of 42 results (offset: 0)

### 1. Scaling Laws Have Scaling Laws
**arXiv**: [2310.12345](https://arxiv.org/abs/2310.12345) | **Authors**: Alice Chen, Bob Liu et al.
**Published**: 2023-10-18 | **Category**: cs.LG
> We investigate the limits of neural scaling laws and show that scaling
> efficiency itself follows a power-law decay as model size grows beyond...

*37 more results available. Use `offset=5` to continue.*

arxiv_get_paper

Input

{ "ids": ["2310.12345", "2301.07758"] }

Output (markdown)

## Scaling Laws Have Scaling Laws
**arXiv ID**: [2310.12345](https://arxiv.org/abs/2310.12345)
**Authors**: Alice Chen, Bob Liu
**Published**: 2023-10-18
**Categories**: cs.LG (Machine Learning)
**DOI**: 10.1234/example
**PDF**: https://arxiv.org/pdf/2310.12345

### Abstract

We investigate the limits of neural scaling laws and show that...

---

## A Second Paper Title
...

Output (json)

[
  {
    "arxivId": "2310.12345",
    "title": "Scaling Laws Have Scaling Laws",
    "authors": [{ "name": "Alice Chen" }, { "name": "Bob Liu" }],
    "published": "2023-10-18T00:00:00Z",
    "updated": "2023-10-20T00:00:00Z",
    "summary": "We investigate the limits of neural scaling laws...",
    "categories": [{ "term": "cs.LG" }],
    "primaryCategory": { "term": "cs.LG" },
    "pdfUrl": "https://arxiv.org/pdf/2310.12345",
    "doi": "10.1234/example",
    "journalRef": null,
    "comment": "15 pages, 8 figures",
    "links": []
  }
]

arxiv_search_by_author

Input

{ "author": "Yann LeCun", "limit": 3 }

Output (markdown)

# arXiv Search Results

Showing 3 of 128 results (offset: 0)

### 1. A Path Towards Autonomous Machine Intelligence
**arXiv**: [2206.07682](https://arxiv.org/abs/2206.07682) | **Authors**: Yann LeCun
**Published**: 2022-06-27 | **Category**: cs.AI
> We propose a modular cognitive architecture for autonomous AI systems...

*125 more results available. Use `offset=3` to continue.*

arxiv_search_by_category

Input

{ "category": "cs.AI", "query": "reasoning", "sort_by": "submittedDate", "limit": 3 }

Output (markdown)

# arXiv Search Results

Showing 3 of 891 results (offset: 0)

### 1. Chain-of-Thought Reasoning in Large Language Models
**arXiv**: [2401.00123](https://arxiv.org/abs/2401.00123) | **Authors**: Wei Zhang, Sara Kim et al.
**Published**: 2024-01-02 | **Category**: cs.AI
> We study emergent reasoning capabilities in LLMs and propose a new...

arxiv_list_categories

Input

{ "limit": 5 }

Output (markdown)

# arXiv Subject Categories

## CS

- **cs.AI** — Artificial Intelligence
- **cs.CL** — Computation and Language (NLP)
- **cs.CV** — Computer Vision and Pattern Recognition
- **cs.LG** — Machine Learning
- **cs.NE** — Neural and Evolutionary Computing

Mock Mode

Run without any API key (returns fixture data of identical shape):

MOCK=true npx @fre4x/arxiv

Deploy

{
  "mcpServers": {
    "arxiv": {
      "command": "npx",
      "args": ["-y", "@fre4x/arxiv"]
    }
  }
}

Development

npm install
npm run dev      # tsx, no build
npm run build    # esbuild → dist/
npm test         # vitest unit tests

License

MIT — WE ARE THE FRE4X.