NCBI E-utilities API
NCBI E-utilities provide programmatic access to the National Center for Biotechnology Information databases, including PubMed, GenBank, Gene, Protein, ClinVar, and SRA.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Public scientific API — no sensitive data returned. API key is a simple query parameter (low security), but data is publicly available anyway. TLS enforced on all endpoints.
⚡ Reliability
Best When
Building biomedical research agents that need to cross-reference literature (PubMed), sequences (GenBank), or clinical variants (ClinVar) from a single authoritative source.
Avoid When
You need to download millions of records in bulk — use NCBI FTP mirrors or Entrez Direct for bulk retrieval rather than the API.
Use Cases
- • Literature mining agents querying PubMed for research papers on a topic and extracting structured metadata
- • Genomics pipelines fetching DNA/protein sequences from GenBank or RefSeq by accession number
- • Clinical research agents querying ClinVar for known disease-associated variants
- • Drug discovery agents linking genes to pathways via NCBI Gene and cross-referencing publications
- • Automated submission monitoring for newly deposited sequences in SRA or nucleotide databases
Not For
- • High-throughput bulk downloads — use NCBI FTP or AWS Open Data snapshots instead
- • Real-time clinical decision support — data can lag by days for new submissions
- • Applications requiring guaranteed SLA — NCBI does not publish formal uptime guarantees
Interface
Authentication
API key optional but strongly recommended. Without key: 3 requests/sec limit. With key: 10 requests/sec. Key registration via NCBI account (free). Key passed as query parameter.
Pricing
Completely free public API funded by NIH. No paid tiers. Commercial use permitted with appropriate rate limit compliance.
Agent Metadata
Known Gotchas
- ⚠ Rate limits enforced per IP — agents must add delays between requests or register an API key
- ⚠ Responses are XML by default; JSON available for some endpoints via retmode=json but not all
- ⚠ ESearch returns UIDs that expire after 8 hours on NCBI history server — agents must use within the session
- ⚠ Large result sets require usehistory=y + WebEnv parameter chaining across ESearch/EFetch calls
- ⚠ Some databases (SRA) have much higher latency and may timeout for large queries
- ⚠ NCBI BLAST API is separate from E-utilities and has its own endpoint and RID polling pattern
Alternatives
Full Evaluation Report
Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for NCBI E-utilities API.
Scores are editorial opinions as of 2026-03-06.