Video and audio transcription · Text and subtitles

Audio and video — to text in a couple of minutes

Upload a file, drop links to clips or capture video straight from an open browser tab. OctaScribe hears it all and turns it into plain text or ready-made subtitles. Fast and accurate speech recognition with no complex settings.

FileClip linkTab captureRecorder
6
AI summary styles
4
speakers in diarization
3 h
length per recording
100%
on our servers
Transcription
Diarization
Whisper AI
Up to 4 speakers
AI summaries
Video player
Timeline
Frame screenshots
PPTX · PDF
Our servers · Russia
Transcription
Diarization
Whisper AI
Up to 4 speakers
AI summaries
Video player
Timeline
Frame screenshots
PPTX · PDF
Our servers · Russia

Ways to upload

Four ways to bring in a recording

No need to convert anything or install extra software — bring your recording however is convenient, and we handle the transcription.

📁
File upload
Drag in audio or video — mp3, wav, m4a, mp4, mov, mkv, webm. Up to 3 GB and up to 3 hours per recording.
Audio and video
🔗
Clip link
Paste a link from YouTube, Vimeo, Telegram, Google Drive and ~1500 more sites — we download and transcribe it for you.
~1500 sites
🖥️
Tab capture
A call running in a neighboring browser tab? We grab the audio and picture right from it — no third-party screen recorders.
Zoom · Meet · any call
🎙️
Microphone recording
Built-in recorder: hit Record, speak — and send it straight to transcription. Nothing to download.
Right in the browser

Demo

What happens after upload

Upload a file — OctaScribe detects speakers, timestamps lines and prepares summaries. Everything runs on our servers, files don't go to third-party services.

🔒
Full privacy
Whisper and pyannote run on our servers — files don't go to outside clouds.
🎬
Video — no conversion
mp4, mov, mkv, webm — upload as is. Audio is extracted automatically.
⚡
Live status stream
Processing progress visible in real time via WebSocket. Leave and come back — the result will be waiting.
OctaScribe — processing
🎙️
Drop a file or click to pick one
Or click to watch the processing demo
MP4MOVMP3WAVM4A+ ещё

Features

Everything you need to work with recordings

🔒
Files stay in-house
Whisper and pyannote run on our servers in Russia. Your audio and video are never sent to outside clouds.
Our servers · Russia ✓
🎬
Video player with timeline
Built-in player with subtitles and colored speaker tracks. Click a line — jump to that point.
Video only
📸
Frame screenshots
Pause — hit "Snap". Frame saved to the gallery. Pick the ones you need and assemble a presentation in one click.
Video only
🎯
Fragment selection
Pick specific lines with checkboxes. AI summaries and the deck use only the selected parts — not the whole recording.
Audio and Video
📊
Slide deck from the meeting
Choose screenshots and fragments — OctaScribe assembles a PPTX with slides, AI titles and bullets. Export to PDF in one click.
PPTX · PDF
🔄
One transcript — all styles
Switch summary styles without re-processing. Every version is saved — you can come back to any of them.
6 styles · History

Summary styles

6 voices for one meeting

One transcript — any style. Prompts live in the database and can be edited without redeploy.

📋
Free
Meeting protocol
Neutral · Official
Participants, agenda, decisions, tasks with owners and deadlines. Ready to send without edits.
⚡
Free
Executive
Directive · Concrete
Key takeaways, decisions, next steps. No fluff — only what matters for the leadership.
🤝
Pro+
Diplomat
Balanced · Smoothed
Positions of each side, agreements and disagreements. Phrasing without blame — perfect for negotiations and HR.
📊
Pro+
Analyst
Structured · Detailed
Facts, numbers, arguments, risks and alternatives. For consultants, analysts and strategic decisions.
📰
Pro+
Journalist
Lively · Narrative
A short retelling in news-note format — who, what, why and the outcome. For PR and marketing teams.
🌱
Max
Coach
Supportive · Growth-oriented
Strengths of the discussion, growth areas and reflection questions. For coaches, mentors and HR partners.

How it works

4 steps from file to result

The whole process is automatic. You upload — we process. The result waits until you're ready.

1
Upload the file
Drop or pick audio (mp3, wav, m4a) or video (mp4, mov, mkv). Up to 3 GB and 3 hours.
2
Transcription
Whisper transcribes speech with timestamps. Then pyannote detects who's talking — up to 4 speakers.
3
Work with the text
Rename speakers, highlight fragments. For video — timeline with subtitles and frame capture.
4
AI summary and export
Pick a style — get a ready summary. Export to TXT, DOCX, SRT. Video adds PPTX and PDF deck.

Reviews

What they say about us

FAQ

The essentials, briefly

Everything runs on our servers in Russia: Whisper does the transcription, pyannote handles the speaker split. Your audio and video don't go to any third-party clouds.

Audio: mp3, wav, m4a, ogg, flac, aac. Video: mp4, mov, mkv, webm, 3gp. Up to 3 GB and up to 3 hours per file. You can also import by link from YouTube, Vimeo, Telegram, Google Drive and ~1500 more sites.

Russian and English, plus automatic language detection. Whisper recognizes speech with time codes; you can turn on a larger, more accurate model if needed.

Up to 4 speakers via pyannote diarization. You can rename speakers in one click, or pick plain-text mode if you don't need the split.

Yes. One transcript — 6 styles: Protocol, Executive, Diplomat, Analyst, Journalist, Coach. The style switches without reprocessing, and all versions are saved.

Yes. One account and one balance work across all lifeisocta apps — OctaScribe, OctaQuip, OctaCut and others. Sign up once.

The lifeisocta ecosystem
One account and balance for every app. lifeisocta.ru →
OctaScribeOctaCutOctaTaskOctaQuip+ ещё