Our Partners & Collaborators
Working alongside leading organisations in News, Media, Data Journalism, Governments, Think Tanks and AI & Web3 policy.
Omniscient AI runs ChatGPT, Perplexity, and Google Gemini simultaneously on any claim — and shows you where each one agrees or disagrees, with live citations. Built at the cutting edge of AI × Web3 to deliver real-world utility.
By signing up, you agree to receive product updates and our newsletter from Omniscient AI.
Working alongside leading organisations in News, Media, Data Journalism, Governments, Think Tanks and AI & Web3 policy.
No switching between tools. No copy-pasting. Omniscient AI runs in your browser and checks facts where you read them.
Select any sentence or paragraph on any web page, or use Full Page Analysis on a complete article. No setup required beyond installing the extension.
ChatGPT, Perplexity (with live web search), and Google Gemini each analyse the claim independently and return their verdict — Supported, Disputed, Mixed, or Opinion.
A consensus score shows how much the AIs agree. Each verdict comes with web sources and citations so you can verify independently. Full Page Analysis scores the entire article 0–100.
10 free checks on signup · No card required · Works on any news site
No single AI is perfect. Omniscient runs all three so you see where they agree — and where they don't.
Provides structured, nuanced reasoning on claims with deep knowledge of world events. Excellent at identifying opinion vs verifiable fact, and at citing its reasoning step by step.
The only AI in our stack with real-time web search. Perplexity actively searches the current web for evidence, making it especially powerful for breaking news and recent events.
Google's latest frontier model brings its vast knowledge graph and search-backed understanding to every claim. Provides a strong independent perspective to balance ChatGPT and Perplexity.
Watch the conversation on AI, countering misinformation and disinformation, and the future of trustless trustworthy information in the age of large language models.
An invitation-only gathering of journalists, media leaders, and AI researchers, hosted by the UK Government and British Council — focused on building resilient AI-powered newsrooms.
41 real-world claims. 3 difficulty tiers. Ground truth sourced from CDC, WHO, PolitiFact, FEVER dataset, Reuters, AP, and NASA. Two models evaluated head-to-head — same claims, same models, different methodology.
| Metric | Baseline | VERIFAID | Δ |
|---|---|---|---|
| Accuracy | 61% | 80% | +19pp |
| F1 Score | 0.59 | 0.80 | +0.21 |
| Precision | 0.63 | 0.82 | +0.19 |
| Recall | 0.58 | 0.79 | +0.21 |
| Avg Confidence | 66 | 84 | +18 |
| Claims Tested | 41 total (same set) | — | |
Credits never expire. No subscription required. Pay only for what you use — or go Pro for unlimited access and source verification.
All payments processed in USDT on Polygon. No card required. · 🔒 Source verification requires Pro subscription
Integrate our triple-AI fact-checking engine directly into your publication, CMS, or app — powered by the same engine used at Omniscient.news.
