Skip to content
Need help?

Choose a transcription model

By default, Borumi transcribes your recordings on your computer. For some languages, you might get better results with another model, and this guide walks you through the different options.

When picking a model, keep a few things in mind:

  • For the best accuracy, use a cloud transcription model. It processes your audio in the cloud, so you need an internet connection.
  • To keep everything on your computer, use a local model.

In general, we recommend keeping Auto selected, because it already picks the best local model for your language. If you want to use another model, follow the steps below.

  1. Click the gear icon at the top right to open Settings.

    The Edit tool with the gear icon highlighted at the top right
  2. Click Editing.

    Settings with Editing highlighted in the sidebar
  3. Under Transcription model, choose a model.

    Editing settings with the Transcription model menu open and highlighted

These models run on your computer, so your recordings are never uploaded to the cloud. That’s great for privacy, but the transcript is usually less accurate than with a cloud model. Auto is the default and supports a wide range of languages. You can find the up-to-date list of models under Transcription model in the Editing settings.

Borumi also lets you transcribe your videos in the cloud, which can give you a higher-quality transcript. These are the models available today.

ElevenLabs Scribe v2 is a cloud model from ElevenLabs that supports more than 90 languages. It uses your own ElevenLabs account, so you need an ElevenLabs API key.

0:00
0:23
  1. In your ElevenLabs account, create an API key with access to Speech to Text.

  2. In Borumi, open Settings, click Integrations, and paste the key under ElevenLabs API Key.

    Integrations settings with Integrations highlighted in the sidebar and the ElevenLabs API Key field highlighted
  3. Click Editing, then choose ElevenLabs Scribe v2 under Transcription model.

    Editing settings with Editing highlighted in the sidebar and ElevenLabs Scribe v2 highlighted in the Transcription model menu