Support & FAQ
The shadow video plays but the karaoke highlight doesn't move.
JumpInto polls YouTube's bridge ~2 times per second and interpolates a 30 Hz karaoke clock on top. If the highlight ever stops moving, tap the row to re-anchor; the highlight should resume from the tapped line. If a row appears stuck across multiple plays, restart the video and report it via the in-app flag button so we can attach the transcript id.
The mic button asks for permission every time.
iOS will re-prompt if the Speech Recognition or Microphone permission was revoked in Settings › JumpInto. Re-enable both. JumpInto needs Speech Recognition to score what you said against the target sentence and Microphone to capture audio.
Writing or Speaking feedback returns an error.
Feedback is generated server-side by the JumpInto Worker. If you see an error, your draft is preserved. Tap Edit answer to tweak the text and Retry. If errors persist for more than a few minutes, please reach out via Contact.
Why does my Speaking attempt show a review screen before scoring?
Audio never leaves your phone. Only the on-device speech-to-text is sent to be scored, so we let you fix any mishears in the transcript first. This keeps the API spend honest and the band score accurate.
I'm anonymous: will I lose my progress if I sign in later?
No. Anonymous progress is stored locally, and when you sign in we migrate your scores and bookmarks to your account so the band history you've built up follows you across devices.
How is my IELTS band computed?
Listening and Reading use a conservative half-band variant of the official Cambridge conversion table. Writing and Speaking are scored by a strict-rubric AI examiner trained against the public band descriptors. Your overall band is the mean of the four per-skill best bands, rounded to the nearest half-band: the same unit IELTS reports on the official Test Report Form.
I want to report incorrect content or a transcript error.
Every shadow video has a flag button in the top action bar. Pick a reason, optionally add a comment, and our team will review within 48 hours. You can also reach us at the address on the Contact page.
System requirements.
JumpInto requires iPhone or iPad running iOS 17 or later. Network access is required for YouTube playback and feedback. Speech recognition works offline as a fallback when the server model is unreachable, but accuracy is best online.