Questions a launch reviewer should not have to guess.

No. Yentl is AI-assisted review software. Its verdicts can be wrong, incomplete, or overconfident, so the interface keeps source context, evidence quality, uncertainty, and challenge paths visible.

You can start from a Chrome tab, YouTube link, audio or video file, live microphone, pasted text, one claim, or a direct media URL. Platform-specific options vary: mobile web supports share/import, while current-tab capture depends on the Chrome extension.

Not for the v1 guest flow. Sessions save locally in the browser first. When Clerk and the Yentl database are configured, signing in also enables account-synced saved sessions.

The library keeps browser-local snapshots for guests and can merge account-synced saves for signed-in users. Clearing site data can remove local saves; deleting a cloud save removes it from the account.

No. Media may be processed by Yentl and its listed subprocessors to transcribe and analyze a source. Saved-session records store the review snapshot locally first and, for signed-in sync, in the account database.

Markers are there to support critical reading. They identify patterns such as loaded language, false dilemmas, or confirmation bias and should point back to the quoted language that triggered them.

Yentl should explain the failure and route you to another path: browser capture, uploaded media, pasted transcript, direct media URL, or opening the original source outside the embedded player.

The v1 public preview has no published paid plan. Responsible partner pilots are handled by contact until real commercial terms exist.