For Agents
Search and retrieve metadata for academic publications, datasets, projects, and organizations from the OpenAIRE scholarly graph.
Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the OpenAIRE Graph API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fgraph.openaire.eu%2Fopenaire" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fgraph.openaire.eu%2Fopenaire" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with OpenAIRE Graph API.
Search research products by keyword, type, or funding stream across publications, datasets, and software
Retrieve a publication or dataset record by its OpenAIRE identifier
List EU-funded projects and resolve their participating organizations
GET STARTED
Use for: Search for open access publications about climate adaptation, Find all research datasets funded by a specific Horizon Europe project, Retrieve a publication by its OpenAIRE id, List the projects that an organization participates in
Not supported: Does not host fulltext content, run scientometric analyses, or manage user accounts - use for read-only scholarly metadata lookup only.
The OpenAIRE Graph API exposes the open scholarly knowledge graph of European research outputs, providing programmatic access to publications, datasets, software, other research products, projects, organizations, and data sources. Both v1 and v2 of the Graph API are reachable from the same base URL, with v2 returning the latest enriched metadata records. Records can be retrieved by identifier, listed by entity type, or related through link endpoints that surface affiliations, funding relationships, and citations.
Discover the data sources that contributed a given research product
Walk relationships between research products, projects, and organizations through the links endpoints
Look up persons associated with publications and datasets
Patterns agents use OpenAIRE Graph API for, with concrete tasks.
★ Open Science Compliance Reporting
Research offices and funders use the OpenAIRE Graph API to confirm that publications and datasets produced by funded projects are openly available and correctly linked to their grants. The API returns the project-to-product relationship records needed for Horizon Europe and national open access reporting.
Retrieve every research product linked to project id corda__h2020::abc123 and flag any without an open access fulltext
Literature Discovery for Systematic Reviews
Systematic review teams use the research products endpoint to pull a comprehensive set of publications and datasets matching a query, including European outputs that may not be well covered in commercial databases. Results include identifiers, titles, authors, and links to fulltexts where available.
Search /v2/researchProducts for keyword 'mRNA vaccine adjuvant' and return the first 100 records with identifiers and abstracts
Institutional Research Profile Building
Universities building public-facing researcher and department profiles enrich their internal data with OpenAIRE Graph records to add affiliations, co-authorship, and project participation that may not be tracked locally. Organizations and persons endpoints provide the linkage needed.
Look up an organization by id, list its associated research products from the last three years, and return co-author counts
Agent-Driven Research Assistant via Jentic
An AI research assistant uses Jentic to discover the right OpenAIRE Graph operation for a researcher's question, load the schema, and execute the call. Because the Graph API is open, no credentials are required, but Jentic still handles request signing, retries, and pagination so the agent stays focused on synthesis.
Search Jentic for 'find openaire research products by keyword' and execute with the user's natural-language topic
18 endpoints — the openaire graph api exposes the open scholarly knowledge graph of european research outputs, providing programmatic access to publications, datasets, software, other research products, projects, organizations, and data sources.
METHOD
PATH
DESCRIPTION
/v2/researchProducts
Search research products in v2
/v2/researchProducts/{id}
Retrieve a research product by id in v2
/v1/researchProducts
Search research products in v1
/v1/researchProducts/links
Walk links between products and other entities
/v1/projects
List funded research projects
/v1/projects/{id}
Retrieve a project by id
/v2/researchProducts
Search research products in v2
/v2/researchProducts/{id}
Retrieve a research product by id in v2
/v1/researchProducts
Search research products in v1
/v1/researchProducts/links
Walk links between products and other entities
/v1/projects
List funded research projects
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the OpenAIRE Graph API by hand means handling its unauthenticated read layer and reconciling the parallel v1 and v2 paths for research products, projects, persons, and organizations yourself. Through Jentic you install once, import the OpenAIRE Graph API from the API Directory, and your agent calls it.
Permission scoping
OpenAIRE Graph is a read-only scholarly metadata service, so limit the agent to the operations it needs, such as searching research products or looking up a project by id. Every operation you allow only reads public metadata, and you leave out any you do not want the agent to call.
Credential isolation
OpenAIRE Graph needs no credential for its public reads, and Jentic One executes the calls directly. No token enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'find publications for a project' or 'look up an organization's research output', and Jentic returns the matching OpenAIRE Graph operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using OpenAIRE Graph API through Jentic.
What authentication does the OpenAIRE Graph API require?
The Graph API endpoints used here are open and do not require an API key. OpenAIRE does offer an OAuth 2.0 token flow for higher-throughput personal tokens, but for standard read access no credentials are needed. Through Jentic this means the operation can be executed without any vault setup.
Can I retrieve every dataset linked to a specific funded project?
Yes. Use /v1/researchProducts/links to find relationship records or call /v1/projects/{id} to identify a project, then filter /v1/researchProducts by the project identifier. Datasets are returned alongside publications and software because the Graph stores all four product types in the same collection.
What are the rate limits for the OpenAIRE Graph API?
Anonymous use is rate-limited per IP at the gateway. OpenAIRE's published guidance is to keep traffic well below one request per second per client and to back off on HTTP 429 responses. For higher volumes, register for a personal token at the OpenAIRE portal, which raises the per-token quota.
How do I search OpenAIRE for publications by topic through Jentic?
Search Jentic for 'search openaire research products', load the v2 researchProducts operation, and execute it with a keyword query and optional filters such as type='publication' or fromDateOfCollection. Pagination is handled with page and size query parameters returned in the response.
What is the difference between the v1 and v2 endpoints?
Both versions are reachable from the same /api.openaire.eu/graph base URL. v2 returns enriched and deduplicated records and is the recommended target for new integrations, while v1 remains available for compatibility and exposes the relations-info link endpoints used to walk between products, projects, and organizations.
Does the Graph API include citation counts?
The records returned include bibliographic identifiers and links to the source data providers, but citation counts come from the BIP! enrichment pipeline and are surfaced as measure entries on a research product when present. Not every record carries citation measures, especially for datasets and software outputs.
Can I limit what my agent is allowed to do with the OpenAIRE Graph API?
Yes. Because you run Jentic One yourself, your own rules decide which OpenAIRE Graph operations and credentials the agent may use. Since this is a read-only scholarly metadata service, you can allow only the operations the agent needs, such as searching research products or retrieving a project by id, and leave out any others. Every operation you permit reads public metadata only, so the agent can never write or alter records.
/v1/projects/{id}
Retrieve a project by id