Source Research
Trace a piece of content back to where it came from, who's behind it, and how credible that source is.
Source Research combines automated web research with crowd and AI analysis to answer a question Multimedia Verification doesn't: not just whether content looks authentic, but where it actually came from.
Given a URL, an account handle, or a claim, Crowdee researches the web for corroborating and contradicting sources, synthesizes a source pointer and category — government, private, or company — with an AI stage, has a crowd panel confirm that synthesis, and produces a final credibility score with a written explanation.
What You Get
- A source pointer — where the content was first published or shared.
- Background on the accounts involved — who's posting it, and under what identity.
- A source category — government, private individual, or company.
- A credibility score with explanation — not just a number, but the reasoning behind it.
Cost
3,200 credits per run — covers the web research stage, AI synthesis, and the fact-checking crowd confirmation stage together.
Where to Go Next
verify-source-credibility— the dedicated pipeline: web research, AI synthesis, crowd confirmation, and a final scorecard.- Web Research — how the web-research stage type works and what it returns.
- Source Research composes naturally with Content Gathering — gathered items convert into input data that source research can investigate directly.
How is this guide?
Multimedia Verification
AI + crowd detection of manipulation, deepfakes, and synthetic content across text, image, audio, and video.
AI Result Evaluation
Crowd panels rate how clear, well-evidenced, and actionable an AI-generated verification result is.