How to download transcription models

The different transcription models will allow you to find the most suitable type and style for transcribing your audio or video files.

Below is a list of the models and the characteristics of each one.

Get to know each transcription model

1. base-multilingual (default)

A standard engine offers a good balance between speed and accuracy for general audio transcription. Language is automatically detected.

2. tiny-multilingual

A nimble engine for basic audio transcription needs, favouring speed over accuracy. Language is automatically detected.

3. tiny-english

English variant of ‘tiny’, providing fast and more precise English audio transcription.

4. base-english

English-adapted ‘base’ variant, providing a balanced English audio transcription solution.

5. small-multilingual

A more refined engine that slows down slightly for improved transcription accuracy. Language is automatically detected.

6. small-english

English-focused ‘small’ variant, sacrificing a bit of speed for better English transcription accuracy.

7. medium-multilingual

An advanced engine, prioritizes precision over speed for superior audio transcription results. Language is automatically detected.

8. medium-english

The English-oriented ‘medium’ variant, offers high accuracy at a moderate speed for English audio transcription.

9. large

The premium engine, delivering the most accurate, albeit slower, audio transcription performance. Language is automatically detected.

Downloading transcription models

Open the application and click on the settings button at the bottom of the sidebar.

At the top, you can see the default model.

In this case, as we have not downloaded any yet, the base – multilingual is active by default.

Choose the model you want to download and click on the download button on the right side.

It will start downloading.

Once downloaded, you can choose it as default by clicking on it.

You can only download one at a time, but once downloaded, you will have it available forever unless you delete it.

Now, you can transcribe your files and choose it as a model. You can also select the most suitable model for each transcription.

Up to here is the guide.

If you still have any doubts, write to us; we will solve them as soon as possible.

Was this helpful?