OpenAlex API
Provides a fully open academic knowledge graph covering 250M+ scholarly works with rich entity relationships across works, authors, institutions, journals, publishers, and research concepts — a free replacement for Microsoft Academic Graph.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Fully public open data with no authentication. No sensitive data is accessible. The mailto polite pool identifier is transmitted in plain text but carries no security risk.
⚡ Reliability
Best When
An agent needs a unified cross-discipline academic knowledge graph with stable entity IDs, open licensing, and rich relationship traversal across works, authors, institutions, and concepts.
Avoid When
You need MeSH-coded biomedical indexing, real-time data, or the specialized clinical database coverage that NCBI provides.
Use Cases
- • Map the citation network around a paper or author to trace the intellectual lineage and influence of a research line
- • Identify all papers affiliated with a specific institution or funding body and aggregate their citation impact metrics
- • Discover related research concepts and their co-occurrence patterns to find unexplored interdisciplinary connections
- • Filter open-access works by license type (CC-BY, gold OA) to build a corpus of freely redistributable full-text content
- • Track an author's publication output, h-index, and institutional affiliations across their full career using stable OpenAlex IDs
Not For
- • Biomedical-specific MeSH term queries — use PubMed for fine-grained clinical/biological taxonomy filtering
- • Real-time preprint indexing — OpenAlex ingests arXiv but with a lag; not suitable for tracking submissions within hours
- • Paid or proprietary journal full-text — metadata and OA links only, no content delivery for paywalled articles
Interface
Authentication
No authentication required for the standard API. Including an email address as the 'mailto' query parameter or in the User-Agent header grants access to the polite pool with higher and more consistent rate limits. A premium tier with SLAs is available for enterprise users.
Pricing
Operated by OurResearch, a non-profit. Core API is free. Data snapshots available for bulk download via AWS S3. Enterprise SLA tier exists but pricing is not publicly listed.
Agent Metadata
Known Gotchas
- ⚠ Cursor pagination uses an opaque next_cursor token that is only valid for a short window — agents that delay between pages may need to restart pagination
- ⚠ The 'filter' parameter syntax uses comma-separated key:value pairs with its own escaping rules, not standard query string encoding — malformed filters silently return empty results
- ⚠ Author disambiguation is probabilistic; the same researcher may have multiple OpenAlex author IDs that have not yet been merged
- ⚠ Concept/topic scores are ML-derived and may be inaccurate for niche or emerging fields; do not treat them as authoritative classifications
- ⚠ The 'select' parameter to limit returned fields is essential for performance but easy to forget, causing unnecessarily large payloads for high-volume agent workflows
Alternatives
Full Evaluation Report
Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for OpenAlex API.
Scores are editorial opinions as of 2026-03-06.