Help Center

Answers about uploading, recording, links, editing, exports, languages, privacy and your account in VoqText.

Zuletzt aktualisiert am 6. November 2026

Diese Seite ist derzeit auf Englisch verfügbar.

Getting started

Upload an audio or video file on the home page or any tool page. Guests can try one short file per time window and read a real excerpt of the result.
Create an account to keep your projects. The confirmation email contains a link; until you confirm, transcription in the workspace is paused.
An eligible plan unlocks the full transcript, editor, exports, source-linked highlights and quote templates. Registration alone does not unlock the result.

Accepted formats: MP3, WAV, M4A, AAC, OGG, Opus, FLAC, WebM audio, MP4, M4V, MOV, WebM and MKV. Files are validated by content, not extension.
Size and duration limits depend on whether you are a guest or a member and on your plan. The uploader shows the current limits before you start.
Large files upload in parts. If a part fails, only that part is retried.
Files without an audio track, or that are corrupt, are rejected during validation with a reason.

The microphone is requested only when you press Start recording. Pause, resume, listen back and discard before uploading.
Recording stops automatically at your duration limit. Keep the tab open while recording.
You confirm that everyone being recorded has been informed and consents.

Paste a direct link to a media file you host and VoqText checks it safely. Platform pages such as YouTube, TikTok and Instagram are not downloaded.
When a link cannot be imported, the page explains why and offers the upload path.

Click Edit, then click a paragraph to change it. Ctrl+Enter saves, Esc cancels. Undo and redo replay through the server, so every state is recorded.
Rename speakers in the side panel; names apply across the transcript and exports. Merge speakers when the detector split one person in two.
Save a checkpoint before large changes. Checkpoints can be restored; the current text is kept in history.
If someone else edits the same transcript, you see a conflict notice and reload; nothing is overwritten silently.

Download as TXT, DOCX, SRT, VTT, JSON or CSV from the Download menu. Files are generated from the current text and delivered through a short-lived private link.
Subtitle exports keep cues readable: characters per line, lines per cue and seconds per cue are bounded, and cues never overlap.
Edited paragraphs use proportional timing in subtitles because word timings from the provider no longer match the text.

The interface is available in 14 languages. Spoken-language support depends on the configured speech provider and is listed on the Languages page with the provider’s own accuracy tiers.
Choose a spoken language in the uploader or leave auto-detect on when the provider supports it.

Media and transcripts are stored privately per workspace. Playback and downloads use short-lived signed links.
Guest previews expire after the retention period shown on the preview page unless you claim them with an account.
Deleting a project removes its media, derived files, transcript, exports and search entries. Backup windows are described in the data retention policy.

“The speech-to-text service is not available”: the provider is not configured or is down; the job did not consume minutes.
“This transcript was changed elsewhere”: reload to pick up the newer version before continuing to edit.
Export failed: try again; if it keeps failing, contact support with the request ID.