← All stations
05 N194°·r61 Live

Transcriber

Transcribe, identify speakers, and subtitle, all automatically.

TypeAI media app
Primary languagePython
InterfaceWeb
StatusLive
Overview

What it is

An AI-powered transcription and speaker-diarization app. Upload audio or video and Transcriber produces an accurate transcript, separates and labels individual speakers, and renders a subtitled MP4 with the names you assign.

The whole flow happens in the browser: upload, review, name the speakers, and download.

Features

What it does

01

Accurate transcription

Transcribe audio and video files with word-level timing.

02

Speaker diarization

Detect and separate individual speakers across the recording.

03

Speaker naming

Assign a name to each speaker and carry it through to the output.

04

Subtitled MP4 export

Burn styled captions into a finished, shareable video.

05

In-browser review

Edit the transcript and fix speaker labels before exporting.

06

Queued processing

Long files process in the background without blocking the UI.

Roadmap

Where it's going

Transcription pipeline

Shipped

Speech-to-text with word-level timestamps.

Diarization & naming

Shipped

Speaker separation and user-assigned speaker names.

Subtitle rendering & MP4 export

Shipped

Styled caption burn-in and finished MP4 output.

Live transcription

Planned

Real-time transcription and diarization for live streams.

Technology

What it's built on

AI
WhisperpyannotePyTorch
Media
FFmpegsubtitle burn-in
Backend
PythonFastAPIRedis queue
Frontend
ReactTypeScript