Evidtrace API

Programmatic access to AI claim verification data, provider credibility scores, and article analysis. Updated daily.

Base URL: https://evidtrace-news.vercel.app/api
All endpoints return JSON. CORS is enabled for all origins. Rate limit: 100 requests/minute.

Endpoints

GET /api/meta

Returns current edition metadata, engine status, aggregate statistics, verdict summary, category breakdown, and a list of all available API endpoints.

GET /api/claims

Query all verified claims across articles. Filter by provider, verdict type, category, model, confidence, or full-text search. Includes evidence summaries and verification notes.

Parameter	Type	Description
provider	string	Filter by provider name (partial match, case-insensitive)
verdict	string	Filter by verdict: Supported, Misleading, Unverifiable, Understated, Contradicted
category	string	Filter by claim or article category
model	string	Filter by AI model name (partial match)
q	string	Full-text search across claim text and evidence
min_confidence	float	Minimum confidence score (0.0 to 1.0)
limit	int	Results per page (default: 50, max: 200)
offset	int	Skip N results for pagination

GET /api/providers

Provider credibility rankings with verification statistics. Sort by credibility score, claims checked, or misleading rate. Filter by minimum score or claim volume.

Parameter	Type	Description
provider	string	Filter by provider name (partial match)
sort	string	Sort field: claims_checked, credibility_score, supported, misleading, supported_pct, misleading_pct
order	string	Sort order: asc or desc (default: desc)
min_score	float	Minimum credibility score
min_claims	int	Minimum claims checked
limit	int	Results per page (default: 50, max: 500)

GET /api/articles

Article-level verification results with claim counts, verdict breakdowns, and scores. Optionally include full claims and verdicts for each article.

Parameter	Type	Description
category	string	Filter by article category (e.g., AI Models, AI Reasoning)
model	string	Filter by model name
min_score	float	Minimum verification score
max_score	float	Maximum verification score
q	string	Search article titles, summaries, and model names
sort	string	Sort: score, title, category, claim_count, verdict_count
include_claims	bool	Set to \"true\" to include full claims and verdicts arrays
limit	int	Results per page (default: 50, max: 200)

Example Usage

curl

# Get all misleading claims from Google
curl "https://evidtrace-news.vercel.app/api/claims?provider=google&verdict=misleading"

# Top 10 providers by credibility score (minimum 20 claims checked)
curl "https://evidtrace-news.vercel.app/api/providers?sort=credibility_score&min_claims=20&limit=10"

# Search articles about Claude with full claim details
curl "https://evidtrace-news.vercel.app/api/articles?q=claude&include_claims=true"

JavaScript

// Fetch provider credibility data
const res = await fetch('https://evidtrace-news.vercel.app/api/providers?min_claims=50');
const data = await res.json();
console.log(data.providers.map(p => `${p.provider}: ${p.credibility_score}/100`));

Python

import requests

# Get all supported claims with high confidence
r = requests.get("https://evidtrace-news.vercel.app/api/claims", params={
    "verdict": "supported",
    "min_confidence": 0.9,
    "limit": 100
})
claims = r.json()["claims"]
print(f"Found {len(claims)} high-confidence supported claims")