| What it actually answers | Is this claim true? Each spoken claim gets a 0-100 credibility score checked against live web sources | Was this made by AI? Probability that an image or text was machine-generated |
| Primary input | Video and audio — YouTube, TikTok, Instagram Reels, podcasts — by URL or via the browser extension while you watch | Still images and text passages, uploaded to the web app or sent through the API |
| AI-media forensic depth | Flags deepfake and AI-generated media signals as one signal alongside the fact-check | Purpose-built for this: per-generator attribution (including 'human artist') plus localized highlighting of which regions look generated |
| Evidence you can audit | Every claim links out to the sources it was checked against, so you can read the evidence yourself | Returns a probability score and a per-generator breakdown — a classifier verdict, not external sourcing (by design) |
| Real-time use while browsing | Chrome and Firefox extensions check claims live as the video plays, and surface relevant viewer comments in sync | API available for programmatic integration; a browser extension is listed on their site as in development |
| Free tier | Free tier available; Basic $3.99/mo, Premium $8.99/mo | Basic image and text classification is free — their site states detection will always be free — with paid tiers adding localized detection, model ID, and API volume |