Transcriptor: live audio to text FREE. Transcription. users statistics

Convert audio to text via the Web Speech API. Completely free!

The extension recognizes speech and displays the transcript directly in the browser sidebar. Supports multiple sessions, multiple languages, live text, copying and exporting to .txt.

Ideal for transcribing calls and creating meeting summaries.

The extension can recognize sound either directly from a browser tab or through a virtual microphone.

-- To recognize audio in the “Sound from Tab” mode, you don’t need to install anything additional: go to the site, open the extension, select the recognition language, and press the “start” button.

---

Recognition in virtual cable mode.

---

The advantage of this mode is that the web speech api continues to work as usual, but in parallel you can use other transcription programs that will also take audio from the virtual microphone and process it. (Whisper, soniox and so on. There are also microphone buttons in Perplexity, ChatGPT, AliceAi, well, in general, you can figure out how to use it yourself. Those who understand, understand😉)

In this mode, the Web Speech API recognizes sound from the microphone, which is specified in the browser as the default microphone for sites. To recognize sound from a call, video or tab, you need to “wrap” it into a virtual microphone using a free program for a virtual cable like VB-CABLE (or similar). This is done once:

1. Download and install VB-CABLE (VB-Audio Virtual Cable). After installation, virtual devices “CABLE Input” (speaker) and “CABLE Output” (microphone) will appear in the system. Restart your computer if the installer asks.

2. Enable listening to the VB-CABLE microphone. Open "Sound Options" → "Sound Control Panel" → "Recording" tab → select "CABLE Output" → "Properties" → "Listen" tab → check "Listen from this device" and select your actual headphones/speakers. This way you will hear the sound that goes into the virtual cable.

3. Select the VB-CABLE virtual speaker in the extension. Open the side panel of the extension, expand the "Speaker-Microphone" block and select "CABLE Input (VB-Audio Virtual Cable)" in the list of output devices. The extension will route the current tab's audio into a virtual cable.

4. In a chrome-like browser, open Settings → Privacy and Security → Microphone. Select your virtual cable (CABLE Output or Virtual Audio Cable, etc.) as the main microphone. You can go directly to this settings section using the link: chrome://settings/content/microphone

5. Click "Start" and allow access to the microphone. The extension will begin to recognize the sound coming through the virtual cable and display text in the panel.

Bottom line: tab sound → virtual speaker VB-CABLE → virtual microphone → speech recognition → text in the panel. At the same time, you continue to hear the sound thanks to the enabled listening (from step 2).

FEATURES ZOOM, CONTOUR.TALK (ktalk), YANDEX.TELEMOST, GOOGLE MEET

In such services, selecting an audio output device through an extension may not work - these applications and sites themselves control which speaker the audio is output to. In this case, select the virtual speaker "CABLE Input (VB-Audio Virtual Cable)" manually in the settings of the application or website itself (section "Sound" / "Speaker"). After this, the extension continues to work correctly and recognizes speech - it is only necessary that the sound be output to the VB-CABLE virtual speaker, and how it is directed there (through the extension or through the application/website itself) does not matter.

Privacy: transcripts and settings are stored locally in the browser. No matter how pretentious it may sound, this is the only extension for live transcription that does not require registration or payment.

https://github.com/sovenov/Transcriptor_chrome