Paste an arXiv ID or abstract URL, get a clean BibTeX entry — straight from the arXiv API.
Enter an arXiv ID and get a ready-to-use BibTeX entry. Fetches metadata directly from arXiv.org.
Machine learning, physics, math, and increasingly economics move on arXiv first — which means half the references in a modern paper are preprints. Paste the arXiv ID (from the abstract page or the PDF header) above, and the converter fetches the authors, title, and year from the arXiv API, formatted as a BibTeX entry ready for your .bib file.
Two habits keep arXiv citations clean. First, when a preprint has been published, cite the journal or conference version — the abstract page lists the DOI when one exists (the DOI converter handles it from there). Second, keep the arXiv ID in the entry so readers can always find the open-access copy.
The modern form (2301.07041), the prefixed form (arxiv:2301.07041), or the full abstract URL (https://arxiv.org/abs/2301.07041). Older-style IDs such as math/0309136 follow the same pattern.
From the arXiv API — the same data shown on the paper's abstract page: authors, title, year, and identifier. Nothing is guessed or generated.
If the paper has since appeared in a journal or conference, most venues prefer you cite the published version — check the arXiv abstract page for a journal reference or DOI, and use the DOI converter for it. Cite the arXiv version when it is the only one, or when you specifically discuss the preprint.
An arXiv ID without a version refers to the latest revision, which is what most citations want. If you need to pin a specific version for reproducibility, note it in the entry after conversion.