Source Tracing Benchmark Leaderboard
Daily testing of AI model source quality — URL validity, domain authority (RSI), citation freshness, and alignment accuracy across 50 research prompts.
Discover which models actually cite real, authoritative sources.
Model Leaderboard
Rolling rankings of AI models by source quality across all daily benchmark runs. Filter by time period and research domain.
| # | Model | |||||
|---|---|---|---|---|---|---|
Performance Insights
Discover source quality trends, compare models across time periods, and find which AI best cites authoritative sources. Data updates as new benchmarks complete.
Source Tracing Intelligence
Recent Daily Results
Click any benchmark run to see per-model source quality details
How It Works
Our automated system tests source quality fairly and consistently across all models every day
Daily at Midnight
50 research prompts across 5 domains are tested daily. Every model's response is analyzed for the URLs and sources it cites.
RSI Authority Scoring
Each cited URL is validated (HEAD request) and scored by the Research Source Intelligence engine — measuring domain authority, source type, and credibility weight.
4-Metric Score
Source Authority Score, URL Validity Rate, Source Freshness, and Citation Alignment combine into a composite RSI score. Rankings update daily across 8 time periods.
Get research with verified, authoritative sources
PromptReports.ai uses RSI scoring to route your research through the model with the best source tracing accuracy for your domain.