Skip to content

DOI to BibTeX

Turn a list of DOIs into BibTeX entries, keep only the fields you want and save them as a .bib file.

Step 1: Paste your DOIs

One per line, or separated by commas, semicolons or spaces. A DOI may be bare, a doi.org link or written after doi:, and may sit in brackets or in a sentence: text that is not a DOI is left out. A repeated DOI, or one already in the bibliography, is skipped.

Separated by

Step 2: Choose the fields to keep

Untick a field to remove it from every entry, the ones already fetched and the ones to come. Fields not listed here, such as address or school, are always kept.

Step 3: Your bibliography

Your BibTeX entries will collect here

Each DOI you fetch adds one entry, in the order you pasted them. Fetch more at any time and they are added to the end, ready to copy or save as references.bib.

How-to guide

How to turn DOIs into BibTeX

Paste one DOI or hundreds and the tool looks each one up at doi.org, adding a BibTeX entry for it to a bibliography you can copy or save as a .bib file.

1

Paste your DOIs

One per line, or separated by commas, semicolons or spaces. A DOI may be bare, a doi.org link or written after doi:, and may sit in brackets or in a sentence: text that is not a DOI is left out. A repeated DOI, or one already in the bibliography, is skipped. Load example adds three to try.

2

Check the separator

Auto-detect handles most lists. If it splits yours wrongly, pick New line, Comma, Semicolon or Space.

3

Fetch

Press Fetch BibTeX. Entries arrive one at a time with a progress bar; any that fail are listed with the reason, and Retry failed tries them again.

4

Choose fields and save

Untick the fields you do not want in step 2, then Copy the bibliography or Save .bib to download references.bib.

How resolution works

DOI content negotiation

Each DOI is requested from doi.org with an Accept: application/x-bibtex header. The DOI system routes to the registering agency (Crossref, DataCite, mEDRA) and returns BibTeX, the same mechanism reference managers use.

Batching, repeats & retry

A DOI repeated in the list, or already fetched since you last cleared the bibliography, is skipped. The rest are fetched one after another with a short pause between requests, and any failures are collected into a report you can retry in one click. A request with no reply after 20 seconds is given up. If a server answers that requests are arriving too fast (status 429), the page waits for as long as the reply asks, or 10 and then 20 seconds when it does not say, and asks for that DOI up to twice more; if it is still refused, the batch stops and the remaining DOIs are listed as not tried. A reply that is not a BibTeX entry, such as a web page or an empty reply, is reported as a failure and not added to the bibliography. Unticked fields are removed in your browser from the returned entries.

Privacy note: formatting and filtering happen in your browser, but resolving a DOI necessarily sends it to doi.org to retrieve the metadata, that network step is how DOI resolution works. No account or tracking is involved.

Frequently asked questions

Does this use a private server?
No STEMKit server is involved. Your browser contacts the public doi.org content-negotiation service directly to resolve each DOI.
Which DOIs are supported?
Any DOI whose registration agency exposes BibTeX via content negotiation, this covers the great majority of Crossref (journal/conference) and many DataCite (dataset) DOIs.
Why did some DOIs fail?
Common causes are a typo, a DOI whose agency doesn't return BibTeX, a transient network error, a server that did not reply within 20 seconds, or one that is limiting requests. Each failed DOI is listed with its reason, naming the server that answered, and all of them can be retried in one click.
Can I remove fields like abstract or url?
Yes. Untick a field in step 2 and it is removed from every entry, in your browser. Fields the panel does not list, such as address or school, are kept.
Are duplicates handled?
Yes. A DOI repeated in the list, or already fetched since the bibliography was last cleared, is skipped rather than fetched again, and the count is reported. Two different DOIs for the same work are not detected; the BibTeX Deduplicator can find those by title.
DOI content negotiation Crossref: content negotiation DOI system

Formatting runs in your browser; DOIs are resolved via the public doi.org service.