FactGuard vs FactVerify

FactGuard checks claims, headlines, or URLs and returns a verdict with reasoning and cited sources. It offers Quick, Thorough, and Private verification modes, with a confidence score from 0 to 100 and analysis of framing and potential bias. The web app limits submissions to 1,000 characters; its free tier uses Quick mode and displays the 30 most recent checks. The free plan costs nothing, billed forever, and includes three checks per day. Pro includes unlimited checks and history, Thorough verification, all modes, shareable verdict cards, and bulk export, but its listed plans are marked coming soon. FactGuard also describes an early-access MCP server and a Fact-Check API with a stated free tier of 100 calls per day; API access and team pricing are under development. The privacy policy says claims sent to the backend are not stored per user on FactGuard's servers. Optional on-device verification is not transmitted to FactGuard or third-party processing providers. FactGuard cautions that automated verdicts can be inaccurate, incomplete, or out of date and advises checking cited sources.

FactVerify is a web-based service that automatically checks claims, viral posts, breaking headlines, articles, webpages and documents. It returns an evidence summary, confidence percentage and linked sources. Its four-stage process identifies a testable proposition, uses primary records, examines context and counter-evidence, then produces a calibrated verdict. The retrieval engine can parse court dockets, statutory texts, peer-reviewed journals and agency datasets; listed primary sources also include SEC filings and government datasets. Confidence percentages describe the coherence and corroboration available when analysis runs, not mathematical certainty. Accounts include five automated checks each month, while the homepage offers one check without an account. Real-time checks are private by default and are not indexed by search engines. Selected checks enter a public library with ClaimReview schema, permanent audit links and freshness tracking. FactVerify notes that automated models can miss temporal nuance, sarcasm, complex statutory references or breaking developments, and says the service is informational rather than legal, medical, financial, tax or investment advice.