Skip to content

quelle

Cite your sources.

Local CLI that fetches academic publication metadata and PDFs from free, open academic sources — OpenAlex, Crossref, Semantic Scholar, arXiv, and Unpaywall — and returns them as normalised JSON. A composable building block: pipe the output into any notes system, reference manager, or research workflow.

The name is German for source — in academic German, "Quelle:" is the word that introduces a bibliographic reference.

Install

pipx install quelle
# or
uv tool install quelle

Both install quelle into its own isolated venv and put it on your $PATH. See Install for the long-form guide, first-run bootstrap, and cross-OS paths.

60-second quickstart

# One-time setup — creates config/data/cache dirs and a commented .env.
quelle init

# Resolve by DOI — OpenAlex + Crossref enrichment chain, cached locally.
quelle fetch 10.1109/83.902291

# By arXiv id, with the PDF downloaded into your data dir.
quelle fetch 1706.03762 --download-pdf

# Or by free-text title, JSON output for piping.
quelle fetch "Attention is all you need" --json

Example output:

$ quelle fetch 1706.03762 --json | jq '.title, .authors[].name, .year'
"Attention Is All You Need"
"Ashish Vaswani"
"Noam Shazeer"
...
2017

Every command accepts --json for machine-readable output. See Commands for the full reference.

What it does

  • Fallback chain over open sources. OpenAlex is primary; Crossref enriches the DOI-authoritative fields; Semantic Scholar fills citation graph gaps; arXiv covers preprints; Unpaywall resolves DOIs to open-access PDF URLs. Fetches merge cleanly — the CLI normalises the shape regardless of which source answered.
  • Three lookup keys. DOI, arXiv id, or free-text title. If you only have a Google Scholar link, paste the paper title instead — Scholar has no API and its ToS prohibits scraping.
  • SQLite cache. Keyed by DOI, arXiv id, OpenAlex id, and normalised title — the second query for the same paper is offline. quelle cache stats / list / show / clear inspect and manage it.
  • PDF download chain. --download-pdf tries OpenAlex → arXiv → Unpaywall in order, with content-type and size validation. arXiv's 1-req-per-3s rate limit is enforced globally via a module-level lock so parallel calls stay polite.
  • Scriptable. All commands take --json, exit codes map to an error hierarchy (1 user error, 2 source error, 3 network, 4 internal), and the whole thing is designed to pipe into notes workflows or reference managers.

Why quelle

You want to grab a paper's metadata without opening a browser, without a paid API key, and without writing source-specific HTTP calls yourself. quelle wraps the free open-access ecosystem behind one command and hands you normalised JSON. Good for personal research notes, Claude skills that need to resolve a paper mid-session, and anyone who wants "I have a DOI → I have a Publication" in one step.

Usage and terms

quelle is intended for personal and academic research use — it queries free public APIs on your behalf, and you are responsible for complying with each upstream's terms of service. Bulk scraping, rehosting downloaded PDFs, and commercial repackaging of the output are not supported use cases. See the README for the per-source licence / rate-limit summary.

Learn more