How to Start Transcribing with EKHOS AI

Updated

EKHOS AI runs transcription locally on your own device, so getting started means creating an account, confirming your email, and installing the desktop app from the Microsoft Store. The site describes three steps: sign up (Free or Premium), click the verification link sent to your inbox, then install EKHOS. Before installing, check the hardware requirements and installation instructions, since the software relies on your machine rather than a cloud service.

What you need before you start

EKHOS is Windows-based: installation is done through the Microsoft Store, so you need a Windows PC that meets the listed hardware requirements. The vendor notes that a properly configured PC matters — one reviewer (Tjerk) mentions that MP3 transcription performs best when it uses the video card's memory, which speeds up processing. If your machine is underpowered or the GPU isn't set up correctly, expect slower transcription.

You also need a working email address for verification, and a decision on Free vs. Premium. The site lists Premium at $9/month (marked "Save 25%"), but the Free tier's limits aren't specified in the available material — check the pricing page before assuming what Free includes.

Step-by-step setup

  1. Create your account. Go to ekhos.ai and choose either "Sign up to Premium" or "Sign up to Free." This determines your usage tier from the start.
  2. Confirm your email. Click the verification link sent to your inbox. The account isn't usable until this is done.
  3. Install the EKHOS application. Install from the Microsoft Store. Before or during this step, review the hardware requirements and the installation instructions linked on the site.
  4. Verify the install works. The expected result is a working desktop app that can transcribe without sending audio to a server. If transcription fails or runs unusually slowly, the hardware/GPU configuration is the first thing to check.

What you can do once installed

  • Transcribe audio and video files, including MP3, with no stated usage cap on the unlimited plan.
  • Speaker identification, so multi-speaker recordings are attributed.
  • Work offline — no cloud connection required for transcription.
  • Live transcription from a microphone and from live speakers.
  • 98 languages supported.
  • Built-in media player for proofreading transcripts against the source audio.

Where privacy fits in

The core design choice is that transcription happens on your device: no cloud, no telemetry, and libraries stored on the local hard drive. That's the reason reviewers cite for choosing it — Ron specifically wanted local transcription without data going to a cloud, and Tjerk valued that information doesn't end up on a third-party server. This matters most if you handle legal, medical, law-enforcement, journalistic, or research material where sending audio to an external service is a problem.

The trade-off is that performance depends on your hardware rather than a vendor's servers, so a weak or misconfigured PC will transcribe more slowly than a cloud service would.

Common snags

  • Skipping email verification — the account won't activate.
  • Ignoring hardware requirements — the app is local-first, so an incompatible GPU or insufficient memory shows up as slow or failed transcription.
  • Assuming Free has no limits — the site advertises "unlimited transcription" alongside a $9/month Premium plan; confirm what the Free tier actually allows on the pricing page.
  • Expecting a browser tool — this is a Microsoft Store desktop install, not a web app.