Private beta · access by licence

Research Data API

Fast, reliable access to scholarly metadata for citation verification, discovery and research tooling. Ask whether a reference is real, and get an answer in milliseconds.

  • Millisecond responses
  • Refreshed weekly
  • Hosted in Sydney, Australia

A reference, as it arrives

Vaswani et al. (2017) “Attention is all u need”, NeurIPS

What comes back

// one verified record, with provenance
{
  "title": "Attention Is All You Need",
  "year": 2017,
  "first_author": "Vaswani",
  "snapshot": "2026-11-02"
}
Illustrative only. Typos, truncation and informal citations still find their record.

What it does

Turn references into records you can trust

Reference lists are full of typos, half-remembered titles and, increasingly, citations that were never real. The Research Data API checks them against curated scholarly metadata, so your tools can tell the difference.

Citation verification

Confirm that a reference points to a real paper, and catch the ones that don’t, including fabricated citations from AI writing tools.

Identifier lookups at scale

Resolve hundreds of identifiers in a single batch request. Check a whole reference list or manuscript queue in one call.

Fuzzy title matching

Find the record behind a messy, truncated or garbled reference. Title search tolerates typos, with optional year and author hints.

Provenance on every response

Each answer states which data snapshot it came from, so results are reproducible and auditable long after the check was run.

How it works

A simple API, built to drop into your workflow

One licence, plain JSON over HTTPS, and answers fast enough to check references as they are typed.

  1. Get a licence

    Your organisation receives its own API token, with a daily request allowance sized to your use.

  2. Send identifiers or titles

    Look up known identifiers in batches, or send free-text titles and let fuzzy matching find the record.

  3. Get verified records back

    Receive structured metadata in milliseconds, stamped with the snapshot date. The data is refreshed every week.

Principles

Designed for trust from the ground up

Privacy-first

We count requests; we don’t record what you look up. Unpublished manuscripts and confidential reviews stay confidential.

Licensed access

Per-organisation API tokens with clear daily allowances. No anonymous scraping, no surprises, and data hosted in Australia.

Built for research integrity

Made by people who work on research impact, for the people who protect the scholarly record: accurate, transparent and reproducible.

Who it’s for

For anyone who needs to know a citation is real

  • Research integrity teams

    Screen submissions and investigations for fabricated or mis-cited references.

  • Publishers and editors

    Validate reference lists at submission, before peer review begins.

  • Libraries

    Resolve patron and catalogue references quickly, even when details are incomplete.

  • Research offices

    Clean and reconcile publication records for reporting and impact assessment.

  • Research tool developers

    Add fast metadata lookups to reference managers, writing tools and dashboards.

  • AI and LLM citation checkers

    Ground generated references in real records and flag hallucinated citations.

Private beta: access by licence

Interested in early access?

We’re working with a small group of organisations during the beta. Tell us what you’d like to check and we’ll be in touch.

Email
info@researchimpact.ai
Website
researchimpact.ai

Helpful to include in your email

  • Your organisation and role
  • What you want to verify or look up
  • Roughly how many lookups a day
  • How you plan to integrate it