Bug fixes and stability improvements • Fixed subtitles edited in a chapter not carrying through to the sequence, which could silently lose your edits • Fixed filler removal cutting into the word that follows, which could drop short words entirely • Fixed SRT-only export from a sequence always failing • Fixed previews not refreshing in sequences that merely reference a clip • Fixed the window freezing for a few seconds at launch • Fixed VoiceOver not reading the verification code field, which made signing in impossible • Fixed parts of the inspector staying in English while the app was set to Japanese • Fixed a misleading message after a failed purchase, and the purchase retrying endlessly • Removed a reference to "Founding Member pricing" that we do not offer

This update fixes two rough edges found after 2.0, and adds a way to take what the app built out of the app. - Telops no longer begin with a stray "。" or "、" carried over from the previous line. - Chapter titles in the chapter box now wrap instead of being cut off, so you can read the whole title. - The sequence Inspector can now copy your chapter and segment structure, and the full transcript. Useful for reviewing an edit outside the app — or for attaching real data to a bug report.

NexClip AI 2.0 is a new generation of the app. One long recording now produces three finished videos, and every step stays yours to edit. ONE VIDEO IN. THREE THINGS OUT. Highlight NexClip AI reads the whole video and assembles a chaptered digest, in the original order. You choose the angle it follows, and the same source can be cut around a different theme whenever you want another one. The runtime comes from what the angle actually supports, not from a fixed template. Short Pick a moment worth posting. It opens on a hook, keeps to a single topic and to the length you set, and is reframed to vertical 9:16 or square 1:1 — the crop follows whoever is speaking. Burned-in subtitles keep it readable without sound. Clip Any single topic as its own video. Pick it from the list. GRAPHICS INSERT Chapter headings, speaker lower thirds, telops and cards are suggested from the sequence you already built. Typeface, colour and motion are set once and applied consistently across every part. Every suggestion is accepted or rejected by you, and you can edit the wording before it goes in. FREEFORM GRAPHICS Graphics Insert suggests. This is the other half — place anything you want, anywhere on the timeline. • Insert a graphics segment at the playhead and drag its ends to cover exactly the stretch you want, across any chapter. • Text, images, rectangles, ellipses and lines, laid out against the output canvas, with drag, resize, rotate and snapping. • Or insert an auto-laid-out plate, then add gradients, silhouettes and icons from the inspector. • Save any design as a reusable part. Editing a reused copy never affects the original. AUTOMATIC AUDIO CLEANING NexClip AI detects dead air, filler words and restarts, and cleans the audio for you. Every cut is non-destructive — undo any of them at any time. • Silence — long pauses between words are cut. Turn this on in Settings. • Fillers — "um", "uh", 「えーと」 and the like. • Restarts — near-verbatim repetitions and false starts. THE FINISHING TOUCHES ARE AUTOMATIC • Subtitles appear on their own. Change the look, or turn them off. • Add titles, logos, or an AI-generated intro/outro image. • Export as MP4 to post anywhere. On the Pro plan you can hand off to Final Cut Pro or Premiere Pro instead. AUTO-BUILD IS JUST THE STARTING POINT Every topic stays editable — reframe, trim to the frame, and clean up fillers exactly how you like. BEFORE YOU UPDATE macOS 26 or later is required. NexClip AI 2.0 will not launch on earlier versions. Apple Silicon. Available in English and 日本語. Your video stays on your Mac. Only the audio is sent, for transcription.

This update brings the biggest jump in clip quality since launch, plus key stability and reliability fixes. Better Topic Clips • Smarter trimming of incomplete openers and trailing fragments • Improved handling of long-form recordings (3+ minutes) — fewer awkward jumps and better topic boundaries • Cleaner output for English interviews and podcasts Stability & Reliability • Fixed an app hang when opening previously imported videos • Empty topic-extraction results now surface as a retry-friendly warning instead of a silent failure • Clearer error messages when an unsupported language is detected or topic extraction fails Transcription • Optional speaker count hint and diarization tuning for more accurate multi-speaker separation Under the Hood • Migrated to the modern async AVFoundation API for more reliable file loading • Resolved Swift Package warnings and concurrency issues Thanks for using NexClip AI. Feedback and bug reports are always welcome at support@cyberseeds.com.