| Primary job | Fact-checks the claims spoken in a video, audio clip, or podcast | Traces where an image came from and where else it appears |
| Spoken claims in video or audio | Transcribes the audio, extracts each verifiable claim, and scores it 0-100 with linked evidence | Not applicable — it matches pixels, not speech |
| Video and podcast input | Paste a YouTube, TikTok, Instagram Reel, or podcast link, or run the Chrome/Firefox extension while you watch | No direct video input — you screenshot a frame and search that single still |
| Image provenance ("have I seen this before?") | Doesn't do it — BSmeter has no reverse-image index | Best-in-class: earliest known appearances, visually similar images, and "About this image" context from Google's index |
| What you get back | A per-claim credibility verdict with the sources linked, plus flagged logical fallacies | A ranked list of matching pages and an AI overview — you read the results and draw the conclusion yourself |
| AI-generated media detection | Flags deepfake and AI-generation signals in the video being analyzed | Google Lens can read SynthID watermarks and answer "is this made with AI?" — a definitive signal on watermarked images, which no heuristic can match |