Streamlined transcription and editing.

Cuts that land between words

Every word Glider transcribes carries its own accurate timing, so edits happen at the word, not somewhere near it. No clipped syllables, no half‑breaths, no stray “um”.

• Strike out to redact and cut: struck words disappear cleanly from the transcript, from playback and from any exported media.
• Remove silent gaps: the silences between words close up, with controls for pause threshold and edge buffers.
• Clip a highlight: strike out everything but the moment you want, and the export starts and ends right on the word.

An hour transcribed in under a minute

Glider's on‑device speech model is fast and accurate. It transcribes and diarizes an hour of audio in less than a minute, timing every word as it goes, with nothing uploaded.

• Speakers, identified: diarization tells voices apart automatically, so the transcript arrives with speaker labels.
• Download once, run locally: the model is fetched a single time, then everything runs on your machine.

Edit, correct, search and replace

Edit the usual way: fix words, add punctuation, split and merge paragraphs. Find and replace across the whole transcript, with the active match highlighted. The timings look after themselves.

Captions, down to the word

Switch into caption mode to review and fine‑tune cues line by line, with timings kept in sync.

• Hear exactly what you're captioning: play clip plays just that cue's stretch of audio, so you can check the words against the voice.
• Jump to in and out points: go straight to where a cue starts or ends, then nudge the timing until it lands.
• Insert, merge and delete: reshape cues in a click, and Glider sorts out the timings around them.
• Karaoke‑style burned‑in captions: bake read‑along captions into the video, with the current word highlighted, Reels‑style.

Readable captions, by the numbers

Every cue shows its reading rate, and Glider flags any caption that goes over your limit, so viewers can keep up.

• Characters per second or words per minute: cps is the streaming convention and wpm the broadcast one. Use whichever your house style does.
• Sensible limits: flag above 17 cps, the usual limit for English, or 180 wpm, the upper end of broadcast practice. Flags are advisory only.
• Line length and breaks: set the characters per line, and start a new caption at each paragraph if you like. A change of speaker always starts a new one.
• Colour‑coded speakers: optionally give each speaker their own colour in exported WebVTT and SRT captions.

Flexible export

Export the whole recording or just your edit, with strikeouts removed and silences skipped. Everything is encoded on your device.

• Speed it up, keep the voice: change the speed or set a target length, and the pitch stays natural.
• Burned‑in captions: bake word‑level, highlighted captions into the video.
• Everything alongside: download an interactive transcript that links the exported media, WebVTT and SRT captions, and a shareable project file with your edits applied.
• Share a highlight: strike out everything but the moment you want, burn in word‑level captions, and export it ready to post. Film in portrait and it stays portrait, ready for Reels, TikTok or Shorts.
• Documents too: export the transcript as plain text, Markdown or Word.

A view that adapts to you

Arrange Glider around the way you work, and turn on paragraph timecodes whenever you need them.

• Compact view: hide the media and sidebar and keep just the transcript and controls, in a window that fits beside everything else.
• Picture‑in‑picture: pop the video out to float wherever you like, or onto a second screen, while you edit on the first.