arXiv Limits Submitters to Two Submissions Per Calendar Month

What Happened

arXiv introduced a submission cap limiting any submitter to two preprints per calendar month, effective across all categories and mirrors. The policy applies at the account level rather than per-author or per-category, meaning the restriction follows the submitting identity. arXiv attributes the change to sustained growth in submission volume and corresponding strain on moderation capacity, which is largely volunteer-run and has not scaled at the same rate as inbound manuscripts.

Why It Matters

The cap converts preprint posting from an unbounded operation into a rate-limited one, and rate limits change institutional behavior. Labs that batch-publish — releasing four to eight papers in a single week to coincide with conference deadlines, benchmark releases, or funding cycles — lose the ability to concentrate output. The constraint is not primarily about arXiv's storage or bandwidth; it is about human review throughput, and that framing matters because it signals which bottleneck the organization believes is binding. For high-volume industrial labs, the cost of exceeding the cap is now coordination overhead: account rotation, staggered scheduling, or migration to alternative venues. For academic groups with a single corresponding author managing all submissions, the cap functions as a hard ceiling on monthly output regardless of research velocity. The policy also creates a measurable asymmetry between institutions that can distribute submission authority across many accounts and those that cannot.

Technical Details

The limit is enforced per submitting account, not per ORCID, affiliation, or IP, which means the practical constraint depends on how a lab provisions arXiv credentials. Two submissions per calendar month is a rolling reset at month boundaries rather than a 30-day window, so a lab can submit two papers on the 31st and two more on the 1st — a detail that will shape scheduling behavior. Endorsement status is unaffected; endorsed submitters retain category privileges but still face the cap. Moderation queues remain the underlying constraint: arXiv's volunteer moderators review flagged submissions, and volume growth has increased both queue depth and false-positive rejection rates. The policy does not appear to include an automated exception path for institutional or bulk submitters, nor a documented appeals mechanism specific to the cap. Cross-listing rules are unchanged, so a single submission can still appear in multiple categories — the cap counts submissions, not listings.

Operational Impact

Research groups that previously released monthly batches now need a submission calendar with named ownership, because the cap is tied to whoever holds the account. Expect a shift toward earlier, smaller releases: instead of one comprehensive paper, labs may split findings across two submissions or hold material for the next month, which lengthens the interval between result and public availability. Internal review processes that gated on "ready for arXiv" now gate on "ready and within quota," introducing a queueing discipline that most labs have never had to maintain. For tooling vendors and internal platforms, this creates demand for submission schedulers, account registries, and pre-print staging systems that track quota consumption across a research organization. Labs with heavy output will evaluate SSRN, HAL, institutional repositories, and direct-to-conference tracks as parallel channels — not replacements, but overflow capacity. The marginal cost of a preprint just went from near-zero to quota-constrained, and quota-constrained resources get allocated deliberately rather than reflexively.