Realtime Source-first AI-Powered

Lets focus on the facts

FactTrace evaluates any quote or snippet against trusted reporting, measures bias and reliability, and returns an instant verdict you can share with your team or community.

  • Cross-references claims against a curated reliability index.
  • Normalizes vague questions into clear statements before scoring.
  • Summarizes why a claim is factual, misleading, or undecided.
Example analysis

Everyone should drink 8 glasses of water a day.

Our sources believe this claim is false.

  • Summary: The claim that everyone should drink exactly 8 glasses (8 oz each) of water daily is not supported by current scientific evidence. The News Medical article explains that the 8‑glass rule originated from a 1941 guideline that was later oversimplified, and modern recommendations from the National Academies suggest total water intake of about 3.7 L for men and 2.7 L for women, including water from food, with needs varying by age, activity, climate, and health.
  • Read the article

Claim workbench

Paste text, get a verdict in seconds.

  • Paste only the relevant passage.
  • We normalize questions into statements automatically.
  • Each verdict includes source reliability and a summary.
0 words • 0 characters

How it works

Three passes for trustworthy answers.

We combine news search, outlet reliability scoring, and a constrained LLM pass to keep verdicts grounded in verifiable reporting.

01

Find reliable coverage

We map your claim to current reporting and filter results against a curated bias & reliability index.

02

Extract the evidence

Articles are scraped, cleaned, and normalized into structured evidence packets for the language model.

03

Score the claim

A deterministic Groq model returns factual vs. misinformation percentages plus a concise summary.