For institutions
For universities, libraries, and departments
Give your researchers a calibrated second opinion before they submit — AI Review, Pre-Check, the Journal Finder and research-gap topics, on volume credits arranged by a person.
What an institution can do with it today
The first four run today, for any account; the last is planned. What an institutional arrangement adds is the credits, not a different product.
- Pre-submission check: researchers run AI Review on a manuscript before it goes to an external journal — 8 specialist agents on the full PDF, one structured report, 30 credits per manuscript.
- Early screening: Pre-Check estimates the chance of acceptance at top-tier, mid-tier and open or emerging journals from a title and abstract (15 credits), and the Journal Finder a ranked shortlist of target journals with predatory-journal flags (15 credits).
- Scoping new work: searching research gaps is free; unlocking one topic — the filled comparison table, the gaps written from it and a proposal draft — is 50 credits, with gaps mined from a 4.5M-paper local library.
- Integration: the review is callable over the API with an account's API key (4 credits per agent call), so a preprint server or an internal workflow can run it on uploads. Any badge or display on your side is yours to build.
- Publication route (planned, not open): Science AI Journal is not accepting manuscripts for publication. Any fee, and whether an institution could cover it, will be stated when the route opens.
How institutional access works
There is no self-serve institutional plan to buy. Tell us about your team on the institutional access form and a person follows up — typically within 1–2 business days — to arrange a volume credit grant and account setup that fits your usage.
The credits spend on the same tools, at the same rates, as any account. Each researcher keeps their own account; a free account starts with 25 welcome credits, and credits never expire. There is no shared institutional dashboard today — if your workflow needs one, say so on the form.
Data posture
We keep this short because it has to be true. Manuscripts uploaded for review are never added to the calibration corpus, which is exclusively public peer reviews from open-review platforms. A saved review is kept with the account that ran it and reached through a private link; to have a manuscript or a whole account deleted, email [email protected] from the account's address.
The Privacy Policy is the full statement — what we collect, which processors see it (Anthropic runs the model inference), and how long it is kept.
How to start a conversation
If you represent a university, library, department, funder or preprint server, use the institutional access form, or email [email protected] with a short note about your team and what you want to run.
We do not run outbound sales calls. We do answer email.