canonical: https://jentic.com/apis/openaire.eu/openaire # OpenAIRE Search API Jentic publishes the only available OpenAPI specification for OpenAIRE Search API, keeping it validated and agent-ready. OpenAIRE is a European open-science infrastructure that aggregates publications, datasets, software, and projects from across institutional repositories, funder databases, and data archives. The Search API exposes six endpoints that let an agent query the OpenAIRE Graph by research-product type - all products, publications, datasets, software, miscellaneous outputs, or research projects - with consistent filtering and paging. It is well suited to literature-review, funding-intelligence, and open-science discovery agents. ## For AI agents Search the OpenAIRE Graph for European research outputs - publications, datasets, software, and projects - through six consistent search endpoints. No API key required. ## Scope Does not provide full-text downloads, citation graph traversal, or impact metrics - use for OpenAIRE Graph metadata search across publications, datasets, software, and projects only. ## Capabilities - Search across all OpenAIRE research-product types in one call - Filter by author, year, keyword, funder, or DOI when querying publications - Search the OpenAIRE-indexed dataset catalogue by title, author, or repository - Find software outputs deposited in OpenAIRE-linked repositories - Pull EU and national funded research-project records by funder, programme, or grant id - Page through OpenAIRE results without managing an API key ## Use cases ### Open-Access Literature Review Build a literature-review pipeline that pulls open-access publications matching a research topic from OpenAIRE. GET /search/publications takes structured filters (keywords, year, author, funder) and returns paged records with DOIs and access URLs, which a research agent can rank and summarise downstream. Example prompt: Call GET /search/publications with keywords='climate adaptation' and fromDateAccepted='2024-01-01' and return the top 50 records with title, authors, and access URL. ### Funded-Project Intelligence Track EU and national research-project funding by programme, funder, or grant id to feed a research-strategy or business-development agent. GET /search/projects returns the structured project record with funder, programme, start and end dates, which an agent can join with publication output via the OpenAIRE Graph. Example prompt: Call GET /search/projects with funder='EC' and fundingStream='H2020' and return projects with end dates in the last 12 months for outcome reporting. ### Dataset and Software Discovery Find research datasets and software linked to a topic or institution so a reproducibility agent can verify or extend a published result. GET /search/datasets and GET /search/software return paged records with the upstream repository and a persistent identifier per item. Example prompt: Call GET /search/datasets with keywords='single cell rna seq' and return the top 20 datasets with their repository and DOI. ### Cross-Type Research Discovery Run a single query across every OpenAIRE research-product type to pull the most relevant outputs regardless of format. GET /search/researchProducts is the canonical endpoint for this and returns a unified result list, which is useful when an agent does not know upfront whether the answer lives in a paper, dataset, or piece of software. Example prompt: Call GET /search/researchProducts with keywords='quantum error correction' and return the top 30 results across publications, datasets, and software. ### AI Agent Research Assistant Expose OpenAIRE Search as a Jentic tool so an LLM research agent can answer questions grounded in real European research outputs. Because the API is keyless, the Jentic flow is just a search and an execute, and the agent can fan out across the six endpoints without juggling credentials. Example prompt: Search Jentic for 'search OpenAIRE for research publications', load the GET /search/publications schema, and execute it with the user's research keywords, returning the top 10 records with DOIs. ## Key endpoints | Method | Path | Description | | --- | --- | --- | | GET | /search/researchProducts | Cross-type search over all OpenAIRE research products | | GET | /search/publications | Search publication records | | GET | /search/datasets | Search research-data records | | GET | /search/software | Search research-software records | | GET | /search/other | Search miscellaneous research products | | GET | /search/projects | Search funded research projects | ## Key resources - **Research Products** — Cross-type search across all OpenAIRE outputs. - **Publications** — Search publication records with author, year, and funder filters. - **Datasets** — Search research-data records linked to OpenAIRE-indexed repositories. - **Software** — Search research-software outputs deposited in linked repositories. - **Other** — Search miscellaneous research-product types. - **Projects** — Search EU and national funded research-project records. ## Why Jentic - **Setup:** Wiring the OpenAIRE Search API by hand means constructing its search query parameters across product types and paging the Graph results yourself. Through Jentic you install once, import the OpenAIRE Search API from the API Directory, and your agent calls it with no credential to manage since the search endpoints are open. - **Permission scoping:** The OpenAIRE Search API is read-only and takes its search terms as query parameters rather than a resource id in the URL path, so limit the agent to the operations it needs, such as searching publications, datasets, or funded projects. You credit the agent only with the operations you allow. - **Credential handling:** The OpenAIRE Search endpoints require no credential, so there is nothing to store; any configuration your Jentic One instance holds is kept encrypted and never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'search research publications on a topic' or 'find funded projects', and Jentic returns the matching OpenAIRE operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **Crossref API** — DOI registration agency with a global publication-metadata search. - **DataCite REST API** — DOI registration agency for datasets with a global research-data search. - **CORE API** — Aggregator of open-access papers with full-text search and download. ## FAQ ### Why is there no official OpenAPI spec for OpenAIRE Search API? OpenAIRE publishes its API documentation on graph.openaire.eu but does not ship a single machine-readable OpenAPI document for the search endpoints. Jentic generates and maintains this spec so AI agents and developers can call OpenAIRE Search API via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer. ### What authentication does the OpenAIRE Search API use? The public search endpoints listed in this spec do not require authentication. Higher-tier features documented on graph.openaire.eu may require a registered token, but GET /search/publications and the other search endpoints in this spec can be called without one. ### Can I search across all OpenAIRE output types in one call? Yes. GET /search/researchProducts returns a unified result list across publications, datasets, software, and other research outputs, which is the right entry point when the agent does not know the format of the relevant material upfront. ### What are the rate limits for the OpenAIRE Search API? OpenAIRE applies fair-use limits on the public search endpoints rather than a strict per-key quota. Plan for occasional 429 responses on bursty queries, page through large result sets rather than asking for huge page sizes, and consult the OpenAIRE Graph documentation for current limits. ### How do I search OpenAIRE for publications through Jentic? Run pip install jentic and search for 'search OpenAIRE for research publications'. Jentic returns the GET /search/publications operation; load the schema and execute with keywords, fromDateAccepted, or other filters to receive paged publication records. ### Can I retrieve the full text of a publication through this API? OpenAIRE returns metadata and access URLs (including DOIs) but does not redistribute publisher full text. Follow the access URL or DOI from the response to retrieve full text from the original repository when available. ### Can I limit what my agent is allowed to do with the OpenAIRE Search API? Yes. Because you run Jentic One yourself, your own rules decide which OpenAIRE operations the agent may call, so you can allow only the searches it needs, such as GET /search/publications or GET /search/projects, and withhold the others like GET /search/datasets, GET /search/software, or the cross-type GET /search/researchProducts. Every OpenAIRE Search endpoint is read-only and takes its search terms as query parameters rather than a resource id in the path, so scoping the agent to a specific operation is the main control you have. This lets you keep a literature-review agent to publication search while a separate funding agent is limited to project lookups.