USER GUIDE / 01

User manual

From first setup to subtitles, meetings, voices and interpretation. Start with the task you need.

Screenshots show the app interface with example settings and sessions. Use your own models, services and recordings.

First time here? Start with a taskFollow illustrated tutorials from installation and model setup to your first result →

Try a short recording first

Before your first session, check that the app captures your voice and displays the transcript. Then enable translation or speech playback. Use the chapters below for setup, subtitles, meetings, voices and interpretation. Screenshot numbers match the instructions below each image; click an image to enlarge it.

Open Setup Wizard, choose an input device and recognition model, and test transcription first.
Record a short passage, stop, and inspect the transcript and meeting record.
Configure and test translation and speech synthesis separately before combining them into interpretation.

Main window: setup, recording and meetings Example

Scenario tutorials · Installation, model setup and task steps Example

Local text cleanup preview · r486 sourceTry explicit Chinese self-correction and scoped dictionary rules locally, compare changes and copy. Live dictation delivery remains disabled; the r481 installer has no entry. Live translation · OpenAISaved connection, automatic input language, translated captions and speech. Windows installation and MSIX certificatesInstall with NSIS, or trust the public test certificate before installing MSIX. Project library & Voices Remembered · r489 source previewSearch original wording across recordings and locate the passage. Generate cited reports or memoirs from entries or a project, review saved affect observations and retain earlier versions. The r481 installer does not include this entry. WelcomeBoundless Flow is a desktop speech workstation with a focus on Chinese-language use.Setup guideFive steps; finish to save, or close to discard unsaved drafts.Transcription & floating subtitlesVerify the input first, then select recognition mode. Observe recording and text status separately.Multilingual translationTest text translation first, recorded subtitles second, and interpretation last.Meetings & minutesAfter recording, select the meeting and review the transcript before creating minutes or exporting.Proofreading & summariesRetain the recognition transcript and treat AI output as suggestions to review.Speech & voicesChoose local models, a model service or a cloud API through one entry, then clone, design and audition.Voiceprints & emotionEnrollment, recording review, independent comparison and calibration. Similarity is not an identity probability.Interpretation workbenchRecognition, translation, synthesis and playback form a live workflow. Try each stage separately before starting continuous interpretation.

LingLu topic minutesFrom a live topic tree to structured minutes and shareable reports.

Troubleshooting & dataLocate the failing stage. Keep original files and the complete error rather than repeating conversion blindly.

Choose a task

For a first installation, read Welcome and Setup. For meetings, read Meetings & Minutes. For model paths or loading errors, read Speech & Voices and Troubleshooting & Data. Every page has a matching language switch. Click screenshots to enlarge them, or print a page to PDF.