What's New in Jumper - June 2026

Jumper v2026.06.28 is now available for macOS (Apple silicon + Intel) and Windows.

Tag collections

Save search results, or whole clips into tag collections with comments, color markers, sorting, and search. Collections can be Personal or Shared; shared ones live in the analysis folder so a whole team can use and update them. You can also filter a people search or visual search to just the clips in a collection.

Read more in the tag collections guide.

Smarter speech analysis

• Multi-channel transcription: A much requested feature that allows you to transcribe one or more selected audio channels, with smart merging that removes duplicate speech when multiple channels carry the same dialogue.
• Speaker detection: Jumper can now identify speakers (diarization) automatically , let you name them, and filter speech search to a specific speaker.

• New models: more accurate speech-to-text for Chinese (20+ dialects), Arabic, Hindi, Cantonese, Korean, Thai, Vietnamese, Filipino, Russian and more. Plus new fine-tuned models for Japanese, Swedish, and Moroccan Arabic/Darija.

Learn more in the speech search docs.

Also new

• Export and import settings: save your Jumper settings and state to a file and load them on another machine. This is handy for teams or when moving to a new computer.
• Notification center: see all the toast messages received in a Jumper session in one place.
• Watch Folder exclusions: Watch Folders can now skip specific subdirectories.

Improvements and fixes

• People grouping has a new ultra-strict regrouping option.
• Fixed visual analysis for videos whose first stream is a cover image instead of the actual video.
• Fixed thumbnail and scrub-frame errors when a scene returns no usable frames.
• Improved reliability when multiple analysis workers write cache files concurrently.

For the full release notes, see the Jumper release notes.

As always, we’d love to hear your feedback. Join us on Discord to share ideas and get the latest updates.