| Unit of analysis | Each individual claim inside the content gets its own 0-100 credibility score with linked evidence | The whole website/domain gets one 0-100 trust score — ratings apply to the source, not to individual articles or videos |
| Video, Reels and podcast content | Built for it — paste a YouTube/TikTok/Instagram Reel/podcast link, or run the extension live while you watch; audio is transcribed and checked | Rates news sources across the web, social, TV/CTV and podcasting platforms, but does not transcribe a given video or check the statements spoken in it |
| Who does the rating | AI transcription and cross-referencing — instant and always-on, but it is a model and it can get things wrong | Trained journalists and editors, with a published nine-criteria methodology, named analysts and an appeals process. Human-vetted and auditable — a real advantage over any AI |
| Coverage of unrated creators | Works on any link, any creator, any topic — no pre-existing rating required | Covers 35,000+ rated news sources (~95% of news engagement in its markets). A random independent creator who isn't a rated news source simply won't carry a label |
| Publisher background and ownership | Doesn't do publisher-level profiles — it evaluates claims, not organizations | This is its core strength. Nutrition Labels explain ownership, financing, editorial standards and track record — context BSmeter does not provide at all |
| Fallacy and synthetic-media flags | Flags logical fallacies and deepfake/AI-generated media signals in the content itself | Tracks AI-generated news sites at the source level, but the extension isn't a per-video deepfake or rhetoric detector |
| Price and access | Free tier; Basic $3.99/mo; Premium $8.99/mo | Roughly comparable, and free in more places than people realize: about $4.95/mo after a two-week trial, but free for Microsoft Edge users and free for schools and libraries |