How to use and assess the directory
On this page
Choose a subfield, identify venues that accept your kind of contribution, and compare the verified free routes. Read the conditions and linked author instructions before preparing a submission. The directory helps build a shortlist; it does not assess an unwritten manuscript or predict its acceptance.
What counts as free#
The default list requires evidence of no mandatory submission or standard publication payment for the recorded route, a peer-reviewed classification, and core editorial standards documented. Read the display policy for the expanded browsing modes. Diamond open access has no author or reader charges. A subscription route may have no author charge while restricting reader access. Selective waivers and institutional funding agreements are conditional, even when they may cover a particular author's costs.
No APC alone does not establish no author cost. Check excess-page charges, required memberships, registration, article types, and the options selected after acceptance. A journal-wide publisher statement has to apply to the recorded journal and route. A blocked page or search excerpt is recorded as limited evidence.
The Diamond Discovery Hub verification guidance checks author-facing charges as well as scholarly evaluation and open licensing. Springer Nature's cost guidance explicitly warns that some subscription journals have page, length, or colour charges. These sources support checking each venue rather than assuming a publishing model makes it free.
Reported metrics#
An acceptance rate is retained only when an official source reports it, with the time period and denominator where available. A rate among externally reviewed manuscripts differs from a rate among all submissions. Historical aggregate rates are not an individual manuscript's probability of acceptance.
Citation indicators describe journal-level citation patterns. Keep the metric name, year, and source with the number, and compare like measures in the same field. They do not establish the quality of a particular paper. This separation follows DORA's guidance on contextual use of indicators. No composite prestige score or worldwide ladder is calculated from missing or incompatible measures.
Evidence and review#
Five distinct agents work on each campaign field: discovery, fees, geography, AI policy, and metrics review. The metrics reviewer also challenges the other findings. Role files are retained under research/, and the build records their hashes. The data include failed verification and excluded leads so a larger list cannot conceal unresolved costs.
Agent review checks the recorded evidence; it is not independent human scientific review, author suitability approval, or endorsement by a venue. Policies can change. Recheck the specific route before every submission and refresh directory records after a material policy change.
Scope and limits#
The covered fields are not an exhaustive taxonomy. Venue coverage is a researched starting collection. Peer-reviewed journals, preprints, and repositories retain separate review labels. Conference venues qualify only if required registration and publication costs are also free; the absence of a conference from the shortlist does not imply it charges a fee.
Author eligibility, submission availability, and restrictions on AI-assisted work can affect suitability even when a route is free. Read the venue's current policies and disclose assistance as required. Do not infer that OSRI's use of Agentic-Powered-Research makes every venue suitable.
AI policy and geography#
The main list requires official, applicable evidence permitting substantive AI-assisted research, analysis or code, and manuscript drafting with human responsibility and disclosure. Permission limited to editing or a silent or inaccessible policy does not establish compatibility. Each entry links its policy and records the disclosure instructions and restrictions; verify their current applicability for the specific work. Where permission follows from a broad assistance rule or combined policy clauses, the record identifies that interpretation and its limits. For Sage venues, conditional partial drafting is interpreted from the combined policy; its prohibition on prompt-generated submissions and analysis of experiences and themes still applies, and stricter journal rules control.
Venue country and region use the owning society or editorial host with a cited basis. Publisher offices, author eligibility and regional research focus are separate fields. An unknown author restriction is not evidence of worldwide eligibility.
Quality evidence#
The quality campaign examines the 146 included AI-compatible venues against a bounded set of public evidence: identity and indexing, peer-review policy and model, named editorial board, ownership, ethics, corrections and retractions, conflicts of interest, complaints and appeals, archiving, licensing, and traceable publication samples. It is informed by the joint Principles of Transparency and Best Practice and is not a formal compliance assessment. Sixteen agent audit tasks and four fresh review tasks are specified in the campaign contract; the primary agent adjudicates the records. This is agent evidence review, not independent human scientific review or endorsement.
“Core standards documented” means the specified core policies, named board, ownership and a traceable sample were documented without a specific medium or high concern. “Verification incomplete” means one or more core checks could not be resolved without such a concern. “Concerns documented” requires a specific evidenced medium or high concern. The default requires core standards documented; incomplete checks are available through an explicit expanded mode and documented concerns remain inspectable under All directory records. These labels are not a score, certification, prestige ranking, or proof that every article was reviewed. Check the linked evidence, access status and limitations for each venue.
Registry-confirmed indexing is distinguished from publisher claims and unresolved checks, with coverage where available. A missing index listing is not misconduct. DOI registration and Google Scholar discovery are not selective citation-index inclusion, and PubMed coverage is not MEDLINE selection. A blocked search cannot establish non-indexing. See the quality review for the scope and interpretation.