HireMindPro logo
← All guides

Voice Clone

Train an AI model of your own voice, then speak through it in real time during interviews and sessions.

Voice Clone lets you train an AI model of your own voice and then speak through it in real time — your microphone audio is replaced with your cloned voice during interviews and sessions. You upload at least 10 voice samples, train a model (charged 10 credits only on success), then select that model and toggle it on. Training runs on HireMindPro's servers; the real-time conversion runs locally via a one-time voice-engine download. Note: the actual feature lives inside the HireMindPro desktop app (not the website), under Settings.

Who it's for
Anyone who wants their microphone audio replaced with an AI clone of their own voice during live interviews or sessions — useful for accessibility, ESL candidates, consistent on-camera delivery, and content/voiceover work. It runs inside the HireMindPro desktop app and requires at least 10 voice samples and 10 credits to train a model.
Cost
10 credits per successful voice clone (one training job). Credits are charged only after training succeeds — failed jobs cost nothing. Uploading samples, downloading the voice engine, selecting a model, and applying your cloned voice in interviews are all free.
Before you start
  • A HireMindPro account, signed in inside the desktop app
  • The HireMindPro desktop app installed (download from /downloads) — voice cloning is not available in the browser
  • At least 10 voice sample files (.wav, .mp3, .m4a, .ogg, .flac, or .webm, each up to 50 MB)
  • At least 10 credits in your account to start training
  • The one-time Voice Engine download (~500 MB) for real-time conversion — the app prompts you to install it
  • Recommended: a microphone routing tool / virtual audio cable so your cloned voice reaches your meeting app, plus a GPU for smoother real-time quality
The Voice Clone page (most voice-clone steps run in the HireMindPro desktop app)
The Voice Clone page (most voice-clone steps run in the HireMindPro desktop app)
1

Open the Voice clone panel

  1. Open the HireMindPro desktop app and sign in.
  2. Go to the Settings / voice area where you'll see three panels: Apply cloned voice, Voice clone, and Hear my voice.
  3. In the Voice clone panel, click the Clone voice button (the one with the + icon). This opens the Voice Clone modal.
Tip: If you only see a marketing page when visiting hiremindpro.com/voice-clone, that's the overview page — the working feature is inside the desktop app.
2

Step 1 — Upload your voice samples

  1. In the Voice Clone modal, find the section labeled '1. Upload Voice Samples'.
  2. Watch the counter (e.g. 0 / 10 files) and the progress bar — you need at least 10 files.
  3. Click 'Upload voice samples' and pick one or more audio files (.wav, .mp3, .m4a, .ogg, .flac, .webm). You can select multiple at once.
  4. Each uploaded file appears in the list with its name and duration. Hover any file and click the trash icon to remove it.
  5. Keep uploading until the counter turns green at 10 / 10 (or more).
Tip: Quality beats quantity: clean recordings from a quiet room with a decent mic produce a far better clone than long, noisy phone audio. Uploading more than 10 good samples improves fidelity.
3

Step 2 — Train your voice model

  1. Scroll to the section labeled '2. Train Voice Model'. It notes: 'Charged ⚡ 10 credits after successful training.'
  2. Type a name for your clone in the 'Enter model name...' field.
  3. Click the Train button. (It stays disabled until you have 10 samples, a name entered, and at least 10 credits.)
  4. Training starts on HireMindPro's servers. You are not charged yet — the 10 credits are only deducted once training completes successfully.
Tip: You can have only one active training job at a time. Starting another while one is PENDING or TRAINING returns 'A training job is already in progress.'
4

Step 3 — Watch training progress

  1. The '3. Training Progress' section lists your jobs with a status badge: PENDING, TRAINING, COMPLETED, or FAILED.
  2. While active, you'll see a live progress bar, a percent-complete readout, and an estimated minutes-remaining figure.
  3. You can close the modal and reopen it later — progress keeps updating automatically.
  4. When it shows COMPLETED you'll see: 'Model ready — select it from the Voice Clone dropdown.'
Tip: If a job shows FAILED, the error message is displayed inline and no credits are charged. Fix the issue (e.g. cleaner samples) and try again.
5

Step 4 — Select and apply your cloned voice

  1. Close the modal. In the Voice clone panel, open the 'Select a cloned voice...' dropdown and pick your finished model.
  2. The first time you use voice features, the app shows 'Voice Engine Required' — click 'Download Now' to install the ~500 MB engine (one-time, runs locally). Wait for the download to finish.
  3. In the 'Apply cloned voice' panel, flip the toggle on. The app syncs your model locally and starts the voice engine ('Syncing cloned voice…', then 'Starting voice engine…', then ready).
  4. Once enabled, your microphone audio is replaced with your selected cloned voice in real time.
Tip: Selecting a different model from the dropdown re-syncs and (if Apply cloned voice is on) restarts the engine with the new voice.
6

Optional — Hear your converted voice

  1. In the 'Hear my voice' panel, toggle the switch on.
  2. This plays the processed (cloned) audio through your default output so you can hear the result.
  3. Turn it off if you don't want to hear yourself looped back during a live call.
Tip: Use this to test and dial in quality before joining a real interview.

Tips & good to know

  • The /voice-clone page on the website is informational only — the real upload/train/apply workflow is in the desktop app's Settings voice panels.
  • You need a minimum of 10 sample files; the Train button won't activate until you hit that count.
  • Credits (10) are charged only after a successful training job — failed jobs are free, so retrying after a bad upload costs nothing.
  • The Voice Engine (~500 MB) downloads once and runs locally; real-time smoothness depends on your computer (a GPU is noticeably better than CPU-only).
  • Cloned audio replaces your mic input — to get it into a meeting app you'll typically route the app's output through a virtual audio cable selected as that meeting's microphone.
  • Each audio file must be 50 MB or smaller; allowed types are .wav, .mp3, .m4a, .ogg, .flac, .webm.
  • Only one training job can run at a time per account.

FAQ

Where do I actually clone my voice — the website or the app?

The voice clone is built into the HireMindPro desktop app, not the website. The /voice-clone page on the site is an overview/marketing page. To upload samples and train a clone, open the desktop app, go to Settings, and find the Voice clone section.

How many voice samples do I need?

At least 10 audio files. The Voice Clone modal shows a progress meter (e.g. 7 / 10 files) and the Train button stays disabled until you reach 10. Accepted formats are .wav, .mp3, .m4a, .ogg, .flac, and .webm, up to 50 MB each.

How much does it cost in credits?

Training a voice clone costs 10 credits. You are only charged after the training job completes successfully — if a job fails, no credits are deducted. The Train button is disabled if your balance is under 10 credits.

How long does training take?

Training runs on HireMindPro's servers. The Training Progress panel shows a live percentage and an estimated minutes-remaining readout. Times vary with server load and hardware; you can close the modal and come back — progress keeps updating.

What is the 'Voice Engine' download and why is it ~500 MB?

The voice engine (RVC) is the local component that converts your microphone audio into your cloned voice in real time. It's a one-time ~500 MB download that runs on your own machine. The app prompts you to install it the first time you enable Apply cloned voice or select a model.

Can I delete a voice clone or a sample?

Yes. In the Voice Clone modal, hover a sample and click the trash icon to remove it. Cloned voice models are scoped to your account and can be removed as well; deleting frees you to retrain or stay private.

My cloned voice sounds rough or robotic — why?

Real-time conversion quality depends heavily on your computer's hardware (a GPU produces much smoother results than CPU-only) and on the quality of your training samples. Use clean, quiet recordings, and upload more than the minimum 10 samples for better fidelity.

Ready to try it yourself?

Open Voice Clone →