FAQ

Sequence is a native Mac app for shaping spoken video and audio from a transcript. Review podcasts and interviews, make reversible text edits, prepare audio, create alternate Variants, work with multiple cameras and reusable assets, then export finished media or continue in Final Cut Pro or DaVinci Resolve. See the feature overview and getting-started guides. Feature coverage reflects the current implementation; check the changelog for availability in your installed version.

Final Cut Pro’s caption tools are useful when you are already editing in its timeline and captions are the remaining task. Sequence starts earlier: read raw footage, shape the rough cut, review repetitive spoken-video cleanup, organize speakers and camera decisions, and prepare audio before continuing in your editor. See the comparison.

Export finished video or audio, TXT or CSV transcripts, SRT, WebVTT, SBV, or ASS subtitle files, still frames, or an editable FCPXML handoff for Final Cut Pro or DaVinci Resolve. You can also share project state in a Sequence Package or add a variant to Assets as a live link.

Handoff fidelity depends on the destination and the composition. Not every editor preserves every camera layout, mix, layer, or title as a native editable feature. Review export warnings and the guides for finished media, Final Cut Pro, and DaVinci Resolve.

Yes, with the optional speaker-detection model installed. Run Detect Speakers after transcription, or enable automatic detection in Transcription Settings. Review speaker assignments, edit names, and filter the transcript by speaker. Detection can be wrong, especially with overlapping speech or difficult recordings.

People provides a local place to manage contributors. Remember Speaker Voices can reuse local voice evidence across projects, but it is not guaranteed identity verification. Those profiles can outlive a project; use the People controls to forget a person when needed. Read the speaker-detection guide.

Yes. Multi-Cam lets you review camera and audio roles, synchronize recordings, and edit programme versions with camera decisions. Supported compositions include picture-in-picture and split screen. For media with multiple embedded streams, you can inspect tracks and choose transcription, playback, and composition sources. Auto Sync requires matching recorded content, and supported media and export behaviour depend on the container and destination. See Multi-Cam editing and importing media.

Yes. Place image, video, and audio assets on the timeline and adjust their timing and placement. A variant added to Assets stays linked, so changes to the original edit can appear where it is reused. This is different from exporting a movie, which is a fixed render. See layers and reusable assets.

Yes. Auto Sync can align a matching external recording using waveform and transcript evidence. Review the result before preparing or exporting the audio. The originals remain untouched, and recordings must share enough matching content. See import and audio setup.

Not currently. You can set framing and crops for individual Variants and adjust supported composition layouts manually. Automatic subject-tracking Smart Crop remains future work. See composition and crop and the roadmap.

Yes. The sq CLI exposes supported project, transcript, media, job, and export actions. A compatible MCP client can find a project, start an editing session, draft a cut, clean up, and export. The default editing workflow creates a protected copy rather than silently replacing your original edit.

Automation is disabled by default. App-backed actions require Sequence to be running, signed in, and unlocked, with the relevant automation access enabled. Your client is responsible for any external AI processing; Sequence does not include a private local language model or an AI-service subscription.

Yes. Sign in to start your seven-day free trial, then download Sequence and sign in with the same account. No card is required.

After sign-in, activation, and model downloads, transcription and editing run locally. Offline access is bounded, not indefinite: direct and Mac App Store licensing use a signed seven-day offline lease that needs periodic online renewal. Distribution-specific subscription requirements still apply. Models, account changes, updates, and iCloud backup need a connection. Media on a network drive must remain reachable. See account and activation.

Buy Sequence directly with one payment for use on up to three Macs, or choose annual or lifetime access through the Mac App Store. Sequence is also coming to Setapp, where it will be included with a Setapp membership. Region-specific direct pricing is shown on the Pricing page. Direct purchases are covered by our Purchase & Refund Policy; Apple and Setapp handle billing and refunds for their own subscriptions and purchases.

Yes. Students, educators, and registered non-profit organizations are eligible for 50% off the standard direct-purchase price.

To apply, get in touch through the Support page.

Students and educators: provide a valid .edu email address, student ID, or institution-issued enrolment or employment letter.

Nonprofits: provide proof of registered charity or non-profit status, such as a charity registration number, 501(c)(3) determination letter, or equivalent.

Once verified, you’ll receive a personalised discount code to use at checkout with your verified email address.

Sequence does not upload your media for transcription. Speech recognition and speaker detection run on-device, and project data is stored in an encrypted local database. Optional iCloud library backup sends the encrypted library to iCloud; it does not back up your original media. The local database does not encrypt original source files or ordinary exports.

If you enable MCP, information returned to your chosen AI client, including requested transcript context, may be processed by that client or its provider. Review its privacy policy before connecting confidential projects. Sharing a Sequence Package or an exported file is also your choice, not an automatic media upload. Account, licensing, and diagnostic traffic are separate from transcription; see the Privacy Policy.

Sequence requires macOS Sequoia 15 or later and an Apple silicon Mac, from M1 onwards. There is no Intel Mac version.

Sequence uses a Core ML conversion of NVIDIA’s Parakeet v3 speech-recognition model on Apple silicon. Speaker detection uses a separate optional Nemotron 3 model pack. Download models through the app before working offline. Model availability and supported languages depend on the installed pack; broader language support is not implied by the interface language. See model setup.

Memory depends on the model pack, recording, and concurrent work. Speech models use several gigabytes while loaded. Sequence schedules speech work and releases idle model memory after its cooldown; the exact memory footprint and timing vary. Speaker detection is a separate processing stage and can add to the job’s resource requirements.

No. The current desktop app requires Apple silicon and macOS. Other platforms remain on the roadmap without a promised release date.

Report an issue through the app or use the Support page. Include the app version and what you were doing. Do not send confidential media or transcripts unless you intend to share them and have permission.

Do not file a public GitHub issue for vulnerabilities. Our security policy requests private reports and coordinated disclosure. Read the Security Acknowledgements page for reporting details.