Skip to main content

Quickstart: Transcribe live audio from your microphone in under 5 minutes!

This article explains how you can transcribe audio in your desired language, live from scratch, on the NeuralSpace Platform in under 5 minutes.


  • Python3.7+
  • Git
  • An activated account on the NeuralSpace Platform
  • A Linux, Windows or a MacOS based system

Step 1: Install NeuralSpace using pip

Install the NeuralSpace command line interface from pip to manage all your services, data, projects and models.

pip install neuralspace

Step 2: Log in to your NeuralSpace Account

You can log in through the terminal using a log in command. This saves your encrypted credentials locally and reuses them for all other APIs/commands.

neuralspace login

This will prompt you to enter the email and hide the protected password.

Step 3: Install the Transcription Service

To use Transcription APIs you will have to install the Service first.

neuralspace install-app -n transcription

This installs the Transcription Service on your account based on the subscription plan you are on.

Step 4: Get the List of Supported Languages

Get supported languages
neuralspace transcription get-languages

This command will display the supported languages with their codes and domains in which they are available.

Step 5: Get the Device ID of your Microphone 🎙️

To start streaming from your microphone you will have to select your microphone's device ID.

Get all available devices
neuralspace transcription list-devices

This command lists the device id of all the I/O ports.

  • The device number with a * mark or a > mark would be your microphone and the device you are looking for.
  • If no devices are listed on running this command, please check your connections to the required devices.

Step 6: Get your Audio Transcriptions

Let's get your first audio transcriptions from the Transcription Service. Make sure to replace the number after -d to your own microphone's device ID.

Get audio transcription
neuralspace transcription stream -L "en" -dom "general" -d 19


  • -L refers to the language code
  • -dom refers to the domain of the model
  • -d refers to the device ID of your microphone.

-L and -dom are to be input by referring to the result of the get-languages command run above.

This command will start catching audio from your microphone and will give you live transcriptions as you speak.

Step 7: Stopping Transcriptions 🛑

To stop the transcription, you can:

  • Press ctrl+c (or cmd+c) to terminate the streaming session.
  • If stream still continues press ctrl+z (or cmd+z) to suspend the session.
  • It is recommended to suspend the session only after first terminating it.

Step 8: Next Steps