ArXiv

An MCP server for searching arXiv, downloading papers, reading bounded full text, retrieving original LaTeX by section, following citation graphs, and maintaining research alerts.

It runs locally over stdio by default. Papers and indexes stay on your machine; search, source retrieval, citation graphs, and downloads call their respective external services.

The command-based integrations requireuv, which providesuvx. Choose your client below; no repository clone or Python environment setup is required.

claude mcp add --transport stdio --scope user arxiv -- uvx arxiv-mcp-server

claude plugin marketplace add blazickjp/arxiv-mcp-server claude plugin install arxiv-mcp-server@arxiv-mcp

codex mcp add arxiv -- uvx arxiv-mcp-server

codex plugin marketplace add blazickjp/arxiv-mcp-server codex plugin add arxiv-mcp-server@arxiv-mcp

hermes mcp add arxiv --command uvx --args arxiv-mcp-server hermes mcp test arxiv

https://github.com/blazickjp/arxiv-mcp-server

macOS users can install a bundled.mcpbextension from thelatest GitHub release:

{ "mcpServers": { "arxiv": { "type": "stdio", "command": "uvx", "args": ["arxiv-mcp-server"] } } }

The default paper directory is~/.arxiv-mcp-server/papers. To choose another directory, append"--storage-path", "/absolute/path/to/papers"toargs.

For older papers that require PDF conversion, run the package with its PDF extra:

{ "mcpServers": { "arxiv": { "type": "stdio", "command": "uvx", "args": [ "--from", "arxiv-mcp-server[pdf]", "arxiv-mcp-server" ] } } }

The supported package is published on PyPI. An unrelated npm package uses the same name, so do not install this server with npm, pnpm, ornpx arxiv-mcp-server.

If an existing installation is missing newer tools

uvxreuses cached tool environments. Force it to resolve the current PyPI release with a supported interpreter, then restart your MCP client:

uvx --python 3.11 --refresh-package arxiv-mcp-server arxiv-mcp-server

If your client still launches an older environment, add"--python", "3.11"before"arxiv-mcp-server"in itsargsarray.

Desktop applications do not always inherit the samePATHas your terminal. Ifuvx arxiv-mcp-serverworks in a terminal but the client reports that the server failed to connect, find the executable's absolute path:

# Windows PowerShell (Get-Command uvx).Source

Replace"command": "uvx"with the returned absolute path, then restart the client. Keep theargsvalue unchanged.

To placearxiv-mcp-serveron yourPATHinstead of launching it throughuvx:

If the command is not immediately available, runuv tool update-shelland restart the terminal. Afterward, use"command": "arxiv-mcp-server"and omit the package name fromargs.

The repository now packages the same MCP server and research skill for both major plugin systems:

Direct MCP installation is the shortest path. Install the plugin when you also want the research workflow that steers the client toward focused searches, bounded reads, citation traversal, and section-level LaTeX retrieval.

Ask your MCP client to callsearch_paperswith:

{ "query": "\"Kolmogorov-Arnold Networks\"", "categories": ["cs.LG", "cs.AI"], "max_results": 5, "sort_by": "date" }

{ "paper_id": "2404.19756" }

{ "paper_id": "2404.19756", "max_chars": 12000 }

Then page through the cached content withread_paper:

{ "paper_id": "2404.19756", "start": 0, "max_chars": 12000 }

Large-content responses includecontent_length,returned_chars,next_start, andis_truncated. Passnext_startinto the next call to continue reading.

{ "paper_id": "1706.03762" }

Get the first page of its section outline withlist_paper_latex_sections:

{ "paper_id": "1706.03762", "start": 0, "max_sections": 100 }

Then callget_paper_latex_sectionusing an ID from that outline:

{ "paper_id": "1706.03762", "section_id": "3.2", "max_chars": 12000 }

LaTeX archives are validated, size-limited, and cached locally before content is returned.

Choose the install variant that matches the features you need:

# Base server uv tool install arxiv-mcp-server # Base server plus PDF conversion uv tool install "arxiv-mcp-server[pdf]" # Base server plus local semantic search uv tool install "arxiv-mcp-server[pro]"

If the base tool is already installed, reinstall the selected variant:

uv tool install --force "arxiv-mcp-server[pdf]"

Thepdfextra installspymupdf4llmandpymupdf-layoutfor papers without usable arXiv HTML. Theproextra adds local embedding dependencies forsemantic_searchandreindex; semantic search only operates on papers already downloaded to the configured storage directory.

The server provides seven MCP prompt workflows. Prompt availability depends on the client; the server provides workflow instructions but does not run a separate model.

For deployments where stdio is not practical:

TRANSPORT=http HOST=127.0.0.1 PORT=8080 \ uvx arxiv-mcp-server --storage-path /absolute/path/to/papers

$env:TRANSPORT = "http" $env:HOST = "127.0.0.1" $env:PORT = "8080" uvx arxiv-mcp-server --storage-path C:\absolute\path\to\papers

{ "mcpServers": { "arxiv": { "type": "http", "url": "http://127.0.0.1:8080/mcp" } } }

Cloud and load-balancer probes should GEThttp://<host>:<port>/healthz. It returns200with bodyokonce the HTTP server is listening. There is no separate/readycheck: if the process is up, it is ready. The stdio transport has no HTTP endpoints.

The server binds to127.0.0.1by default and enables MCP DNS-rebinding protection. If a reverse proxy exposes the server, keep the process on a private interface and provide authentication and network controls upstream. UseALLOWED_HOSTSandALLOWED_ORIGINSfor the host and origin values forwarded by the proxy.

Environment variable names are case-insensitive through Pydantic settings.--storage-pathis a command-line option rather than an environment setting.

Paper text and LaTeX are untrusted external content. A paper can contain text intended to manipulate an AI client into ignoring its instructions or calling unrelated tools.

- Do not treat instructions found inside a paper as trusted commands.
- Use client approval controls for shell, browser, filesystem, and messaging tools.
- Review generated summaries before taking external actions.
- Keep Streamable HTTP private unless authentication is provided upstream.

SeeSECURITY.mdfor the reporting policy and threat details.

git clone https://github.com/blazickjp/arxiv-mcp-server.git cd arxiv-mcp-server uv sync --extra test --extra dev uv run pytest uv run black --check .

{ "mcpServers": { "arxiv-dev": { "command": "uv", "args": [ "--directory", "/absolute/path/to/arxiv-mcp-server", "run", "arxiv-mcp-server" ] } } }

Contributions are welcome. ReadCONTRIBUTING.mdbefore opening a pull request, and useGitHub Issuesfor reproducible bugs or scoped feature proposals.

Anonymous, read-only, source-backed Buddhist scripture search, passage guidance, explanation, and one-time practice planning through four production MCP tools.