← Home
About

We built a radar for the truth.

Misinformation spreads faster than ever. We built TruthRadar to give anyone — not just journalists or researchers — the ability to verify a claim in seconds.


Why We Built This

The gap no one was filling.

Every day, hundreds of millions of people encounter claims that feel true, spread like wildfire, and turn out to be false or wildly out of context. A video "goes viral" before anyone has checked whether the story behind it is real. A screenshot of a statistic circulates for years after the number was debunked. A satire article gets shared as news.

Traditional fact-checking organizations do important work, but they can only cover so much. They focus on politics and breaking news. They can't keep up with the volume of viral content across TikTok, Instagram, Reddit, and X. And their findings are often buried where people who've already seen the false claim will never find them.

TruthRadar exists to fill that gap. It's fast, it covers everything — from paranormal claims to pop culture to health myths — and it publishes every fact-check as a permanent, searchable page.


How It Works

Four steps. Seconds to a verdict.

01

Paste any link

Drop in a URL from TikTok, Instagram, X, Facebook, YouTube, Reddit, or any news article. TruthRadar extracts the claim automatically.

02

AI searches live sources

Our system scans the web in real time — cross-referencing scientific literature, government data, news archives, and expert consensus — not a frozen database.

03

You get a verdict

Within seconds you receive a verdict, a confidence score, a plain-English summary, and the sources used to reach the conclusion.

04

A permanent record is created

Every fact-check is published as a citable page so the work isn't lost. Others searching the same claim find the answer instantly.


Our Verdict System

Four verdicts. Deliberately simple.

Every claim receives one of four verdicts. We deliberately keep the scale simple — complexity in ratings tends to obscure rather than clarify.

TRUE

The core claim is accurate and supported by credible primary sources.

FALSE

The claim directly contradicts verifiable evidence or established fact.

MISLEADING

The claim contains accurate elements but creates a false impression through omission, framing, or lack of context.

? UNVERIFIED

Insufficient credible evidence exists to confirm or deny the claim at this time.


The Technology

Real-time AI. Live sources. No frozen datasets.

TruthRadar is built on Perplexity's Sonar Pro API, which combines large language model reasoning with real-time web search. Unlike AI systems that rely on a fixed training cutoff, Sonar searches the live web when answering — so the fact-checks reflect what sources actually say today, not what the model was trained on two years ago.

Our system extracts the primary claim from whatever content you submit, searches for corroborating and contradicting evidence across credible sources, evaluates the weight of that evidence, and returns a structured verdict. The sources used are always disclosed so you can read the underlying material yourself.

We also maintain a growing library of evergreen fact-checks — the most commonly searched claims about history, science, health, paranormal topics, pop culture, and more. These are written, reviewed, and published as standalone reference pages.


What We Cover

Everything. Not just politics.

TruthRadar covers the full range of claims that circulate online. This breadth is intentional. Misinformation isn't limited to election season. People deserve accurate answers whether they're wondering if a cryptid is real, whether a food product contains what its label says, or whether a historical event happened the way a meme claims it did.

Health mythsScience misconceptionsHistorical claimsParanormal & cryptidsFood factsPop cultureReligionEntertainmentViral social mediaPolitics

Our Limitations

We're transparent about what AI can't do.

AI fact-checking is useful but not perfect. We encourage you to read the sources we cite, not just the verdict. The goal of TruthRadar is to give you a fast, well-sourced starting point — not to replace your own critical thinking.

AI is useful but not perfect

Complex political or scientific claims often require human judgment that no AI system fully replicates.

Honest uncertainty is a valid answer

Highly nuanced claims where experts genuinely disagree may receive an UNVERIFIED verdict — not because the AI failed, but because honest uncertainty is the correct answer.

Real-time events need time

Breaking events may not yet have sufficient sourcing for a confident verdict. We surface what exists and flag when evidence is thin.

Free to Use

Try it right now.

Paste any social media post, article, or link and get a verdict in seconds.