SourceScore
New · v0.1VERITAS claim catalog API — curated, sourced AI/ML records for evidence retrieval. Free public access · no auth.
Methodology v0.1 · 130 sources scored · 384 claim records

How citable is any source in the AI era?

SourceScore checks any URL against an index built from three published citation-quality signals: how rigorously the source cites others, how fit it is as a modern citation, and how often tier-1 publications cite it. Indexed sources return a grade; unknown sources are labeled honestly.

Matches against 130 hand-scored sources — unknown domains show how to get scored.

Composite
SourceScore Index
Composite weighted grade across all three sub-scores.
Score
Citation Discipline
How rigorously a source cites its own evidence.
Reference
Modern Reference
Fitness as a citation in 2026+ AI-era writing.
Tracker
Citation Velocity
How often tier-1 sources cite this URL per week.

Top 5 by SourceScore Index

See all 130 sources →
  1. #1A+·96
    U.S. Securities and Exchange Commission
    sec.gov
    Primary-source regulator publishing every public-company filing (13F, 10-K, 8-K, etc.) since 1934.
  2. #2A+·95
    U.S. National Institutes of Health
    nih.gov
    U.S. federal medical research agency operating PubMed, NCBI, MedlinePlus, and trial registries.
  3. #3A+·95
    DOI (CrossRef Resolver)
    doi.org
    International standard identifier resolver for academic citations (~150M+ DOIs).
  4. #4A+·95
    Federal Reserve System
    federalreserve.gov
    U.S. central bank; primary source for monetary policy + economic data + financial-system statistics.
  5. #5A·94
    Wikipedia (English)
    en.wikipedia.org
    Crowd-edited encyclopedia with ~7M articles and per-article inline citation discipline.

Top 25 sources

The 25 highest-scoring of 130 hand-scored sources. Each row links to the full breakdown with the underlying signals you can re-derive — browse all 130 sources.

SourceIndex
U.S. Securities and Exchange Commission
sec.gov
A+·96
U.S. National Institutes of Health
nih.gov
A+·95
DOI (CrossRef Resolver)
doi.org
A+·95
Federal Reserve System
federalreserve.gov
A+·95
Wikipedia (English)
en.wikipedia.org
A·94
PubMed
pubmed.ncbi.nlm.nih.gov
A·94
U.S. Census Bureau
census.gov
A·94
U.S. Bureau of Labor Statistics
bls.gov
A·94
U.S. Food and Drug Administration
fda.gov
A·94
U.S. Centers for Disease Control and Prevention
cdc.gov
A·94
U.S. National Oceanic and Atmospheric Administration
noaa.gov
A·93
European Central Bank
ecb.europa.eu
A·93
NASA
nasa.gov
A·93
MDN Web Docs
developer.mozilla.org
A·93
European Commission
ec.europa.eu
A·92
Bank of England
bankofengland.co.uk
A·92
U.S. Energy Information Administration
eia.gov
A·92
CERN
home.cern
A·92
PNAS
pnas.org
A·92
European Medicines Agency
ema.europa.eu
A·91
U.S. Patent and Trademark Office
uspto.gov
A·91
U.S. Department of Agriculture
usda.gov
A·91
U.S. Geological Survey
usgs.gov
A·91
OECD
oecd.org
A·91
FRED (Federal Reserve Economic Data)
fred.stlouisfed.org
A·91
New · v0.1 · API in beta

Verified claims for grounded LLM retrieval

VERITAS publishes curated, sourced claim records about AI/ML research through a developer API. Every claim cites primary evidence; 368 of the current 384 claims have two or more sources. Records include SourceScore-issued HMAC integrity metadata and a stable JSON envelope. Use a returned match as a candidate for evidence review—not as an automated truth verdict.

Catalog
Browse claims →
Indexed catalog of verified facts.
Docs
API docs →
curl + JS + Python examples for every endpoint.
Pricing
Pricing →
Free public API · proposed higher-volume offers are a demand test.
How it’s different from a search engine

Search engines optimize for ranking documents. VERITAS optimizes for verifying specific atomic claims: subject + predicate + object + sources + signed envelope. Built for the “is X true and what’s the citation?” problem LLM apps hit at every retrieval step.

Claim records have a stable id (16-hex-char hash over canonical fields), SourceScore-issued HMAC integrity metadata, and a CC-BY 4.0 license. The HMAC tag is not a publicly independently verifiable signature; use HTTPS and the canonical record when checking a published claim.

What does “AI-Citation Quality” mean?

When ChatGPT, Claude, or Perplexity answer a question, they pull from a small subset of sources they consider trustworthy. Two factors decide whether a source makes that subset: citation discipline (does the source rigorously cite its own evidence?) and modern reference fitness (is the source structured for machine retrieval — schema markup, freshness signals, machine-readable archives?).

SourceScore measures both, plus citation velocity (how often the source is cited by other tier-1 sources per week). Together these three sub-scores compose the SourceScore Index — a single 0–100 grade per source.

We ship 130 hand-scored sources at this stage of v0.1. Methodology is intentionally transparent: every score has explicit signals you can re-derive. The production index will expand to 10,000+ sources via the same methodology, with weekly velocity refreshes and quarterly discipline re-audits.

Read the full methodology →How to tell if a source is reliable →About SourceScore
Sister tool · CitationDesk

Is your own site ready for AI citations?

You just checked how citable sources are. CitationDesk audits the page signals on your site that help AI engines retrieve and cite it, then shows the highest-leverage fix. It is a readiness audit, not a claim that an AI engine already cites you.

Get your free AI Visibility Score →