Upload audio or video — HancVoice recognises speech and returns text with timecodes. Speaker labels, speaker names and a summary come with the Business plan; the result can be exported as SRT or VTT subtitles. Names are an AI suggestion (about 80% accuracy) and are easy to fix manually.
Audio or video — an interview, a meeting, a podcast, a lecture. The audio track is extracted from video.
Turn on “Improve with AI” — gently or with text cleanup. Speaker labels, “Detect names with AI” and a summary come with the Business plan.
Download the finished transcript with timecodes and speakers as text or as SRT/VTT subtitles.
HancVoice recognises speech, marks who is speaking and turns the transcript into a readable form. German and English deliver the highest accuracy; many other languages are supported.
From the Business plan: labelling by speaker, and “Detect names with AI” tries to insert names from context (about 80% accuracy). Uncertain ones remain as “Speaker 1”, “Speaker 2” — a name is easy to fix manually.
Two modes: “Gently” — only punctuation and obvious slips; “Clean up text” — a coherent, readable form without filler words.
Export to standard subtitle formats with timecodes — for video, players and editing.
Ready SRT/VTT for videos, lectures and courses — connect them to your player or editor.
Text of calls and meetings with speakers and timecodes — quickly find who said what.
Transcription of interviews and episodes for articles, posts and text versions of content.
The transcription limit is the number of audio minutes per month on your plan. Transcription is available from the Pro plan and above; speakers and summary — from the Business plan.
—
transcription from Pro
100
min/mo
300
min/mo
800
min/mo
Audio and video. From a video file we extract the audio track and recognise speech. The allowed size and length depend on your plan.
Yes, from the Business plan and above. Turn on speaker labelling — the text is split by speaker. The “Detect names with AI” option tries to insert names from context: about 80% accuracy, uncertain ones remain as “Speaker 1”, “Speaker 2”, and any name can be fixed manually. On the Pro plan the transcript is continuous text with timecodes, without speaker labelling.
“Gently” fixes only punctuation and obvious slips, keeping the speech close to the original. “Clean up text” produces a coherent, readable version without filler words and repetitions.
Text with timecodes and speakers, plus subtitles in SRT and VTT formats for video and editing.
The engine is most accurate with German and English; many other languages are supported as well — including automatic language detection.
Upload audio or video right in your browser — transcription is available from the Pro plan.
Transcribe a recording →