Skip to main content
This week we shipped the biggest release of the year: FME V20, our next-generation analysis engine. Alongside it, you’ll find a new pricing page, a press release builder, an upgraded Chrome extension, and a redesigned dashboard with credits and tier visibility.

​
New features

​
FME V20 analysis engine (v4.0.0)

The Forensic Media Evaluation framework has been rebuilt end-to-end. V20 introduces a universal analysis pipeline, a Readability.js-based content filter for cleaner article extraction, and tighter calibration across emotion, ideology, and logical fallacy detection. Higher-tier accounts are automatically routed to V20; free and anonymous tiers continue on V19. Learn more in Core Concepts → FME Framework and the 15 parameters.

​
Public pricing page

A new pricing page lays out every tier, scan quota, and feature side-by-side, plus a credit top-up option for one-off bursts of analysis. See Credits & billing for how credits work inside the dashboard.

​
Press release builder

The new /press page includes a Build Prompt button that generates AI-assisted press release drafts grounded in your scan results.

​
Chrome extension 3.5.0

The extension now:
  • Reflects your tier and remaining quota in the popup.
  • Gates PDF export behind eligible tiers.
  • Syncs document-level emotion (Plutchik-8) and updated V19/V20 fields with the dashboard.
Install or update via the Chrome extension guide.

​
Dashboard: credits, tier badge, quota & transaction history

The dashboard now surfaces:
  • A live credit balance widget.
  • A tier badge showing your current plan.
  • A quota display for the current billing period.
  • Transaction history for top-ups and subscription events.
Details in Dashboard overview and Credits & billing.

​
PDF export

A new PDF export pipeline backs the extension’s “Download report” action for eligible tiers — a portable, shareable version of any scan.

​
Uptime status page

A status page link is now available in the site footer so you can check service health at a glance.

​
Updates

  • V20 latency improvements. Average end-to-end analysis time is materially lower than V19, with a public benchmark report on the test-results page.
  • Pricing tiers corrected. Anonymous and free tiers are now separated, with accurate scan quotas across every plan and a clear credit top-up section. See Credits & billing.
  • Guardrails evaluation results. A new public evaluation tab documents how V20 performs against safety and accuracy guardrails.
  • Homepage refresh. Added a competitor comparison table, customer testimonials, and case-study highlights.
  • V19 ideology calibration. Three calibration patches improve ideology classification accuracy for accounts still on V19.
  • Methodology pages updated to reflect V20 across the public site, About page, and AI-discovery surfaces.

​
Bug fixes

  • Fixed quota display, error messaging, and several false-positive triggers in the analysis pipeline.
  • Fixed subscription checkout to handle the redirect response correctly.
  • V20 defect sweep: emotion-scale normalization, Depth-of-Knowledge calibration, missing database fields, and filter edge cases.
  • Fixed an issue where /press, /pricing, and /intelligence-brief were not reachable without an account.
  • Dashboard scan-count query corrected so daily totals render reliably.
  • Fixed a modal scroll bug that prevented users from scrolling long prompts.
Questions or something not behaving as expected? Check the FAQ or troubleshooting guides.