Ferment — Speech & captions

The speech track

Three ways to put a voice on it:

Record — the voice recorder, with its teleprompter and live waveform.
Extract — the speech lifted out of a video clip.
Synthesize — your script, spoken in a stock voice (free on every plan) or your own cloned one.

Transcripts

Everything spoken gets transcribed on your phone — word by word, with timings, in 26 languages, offline. The transcript is what makes speech editable: it’s how Cuts finds the dead air, the restarts, the sentence you actually meant to keep.

The honest edges: the 26 are European languages plus Russian — no Asian-language transcription on the phone today — and there are no speaker labels; the transcript knows every word and when, not who said it.

Captions

Captions come straight from the transcript, timed to the word — with per-word highlighting if you want it. Pick a style and a layout, apply it everywhere with one switch, and they stay in sync when the edit changes.

Caption panels — the same line in two styles and two layouts, per-word highlight mid-sentence

Cleaning up a voice

Two things, both included on every plan, no credits:

Denoise — a per-clip toggle that strips background noise on-device. Flip it while playing to A/B.
Leveling — Cuts evens out loudness so quiet sentences and loud ones sit together.

Dubbing

Cuts can dub a clip into another language — 23 to choose from. Start from the Dub button on the project bar, or just ask in chat. Dubbing is cloud generation, so it needs a paid plan. The flow is deliberately two-step:

Cuts translates the transcript and synthesizes the voice, timed to the original phrasing. You hear a preview first.
If you like it — and after confirming the cost — Cuts renders the lip-synced video and swaps it onto your timeline. Captions re-derive in the new language automatically.

The original clip is never touched. Re-dub as many times as you like; it always starts from the source.

Dub, two-step — the audio preview playing in chat, then the cost confirm waiting for a yes

Your voice

On paid plans, Cuts can clone your voice into your brand kit: it opens the recorder with a script in the teleprompter, you read it through, and from then on narration and dubs can sound like you. Cloning requires your explicit consent — the terms are short and worth reading: voice consent.