| Core question answered | "Is what they're saying true?" — claim-level accuracy | "Is this footage real, original, and in context?" — media provenance |
| Transcription + spoken-claim fact-checking | Transcribes audio, extracts every verifiable claim, cross-references live web sources, returns a 0-100 credibility score with evidence linked | No transcription or claim-checking layer — the plugin analyzes pixels and metadata, not the argument being made |
| Reverse image search on video keyframes | Not offered — BSmeter cannot tell you if footage is recycled from an older event | Segments videos into keyframes and pushes them to Google Lens, Yandex, Bing, TinEye and Baidu in a couple of clicks. This is its signature capability. |
| Forensic image analysis (EXIF, tampering filters, magnifier, C2PA) | None — no metadata reading or pixel-level forensics | Video/photo metadata extraction, eight image-tampering filters, a deep-zoom magnifier for reading small details, OCR, and C2PA provenance checks |
| Deepfake / AI-generated media detection | Flags deepfake and AI-generation signals inline as part of the credibility report | Dedicated research-grade detectors — a deepfake video tab running multiple models, synthetic-image probability, and a Hiya-backed voice-clone detector (some features require free registration) |
| Output format and expertise required | A finished verdict: scored claims, linked evidence, flagged fallacies, and synced viewer comments. Three modes — Roast, Analyze, Learn. | Raw investigative evidence — search hits, metadata dumps, filter renders — that a trained analyst interprets to reach a conclusion |
| Price | Free tier; Basic $3.99/mo; Premium $8.99/mo | Completely free and open source (MIT), publicly funded — no paid tier at all |