tovrin Start for $199

Voice-over tracks for e-learning, training and explainer studios

Send the script.
Get the finished track.

Tovrin turns your scripts into finished AI-synthesised narration you can drop straight onto the timeline: every word in order, every reading choice listed, mastered and delivered within 2 business days.

  • 48 kHz WAV + MP3
  • −16 LUFS loudness
  • 2 business days
  • No contract
TRACK 01 · osha-ladders.wav 0:00 / 6:13

In the U.S., hundreds of construction workers die every year while on the job, with over 700 fatalities just in the year 2011.

Format
48 kHz · 24-bit · mono
Loudness
−16.0 LUFS · −2.8 dBTP
Length
6:13 · 912 words
Readings listed
10 · 2 questions back
A real delivery, made from an OSHA training transcript (US government work, public domain). Stock AI voice.

You already have the script. What you don't have is a week to book talent, chase retakes, fix "O-S-H-A" in post, and level every file before it fits the edit.

Tovrin does the part between the script and the timeline: we render it in a stock AI voice, read every number and acronym the way a narrator would, master it to broadcast-style loudness, check it word by word against your script, and send you files that fit.

Samples

Three real deliveries. Press play, follow the script.

These are complete tracks exactly as a buyer would receive them, made from public-domain US government scripts of the kind our customers send: a safety training module, an e-learning explainer and an organisation overview. Click any paragraph to jump to it.

Electrocution: Work Safely with Ladders Near Power Lines

Source: OSHA construction v-Tools transcript · voice: American English “Heart” · 0:00 / 6:13

MP3 ↓

Scripts: OSHA (US Department of Labor), NASA Goddard Space Flight Center and NIST, used as US government works. These tracks are our own samples; none of these agencies endorses or is affiliated with Tovrin. Every track on this page is AI-synthesised narration in a stock voice.

Voices

Pick a stock voice. Keep it for every module.

The same line from the OSHA script, read by each voice on our shelf. Your course keeps one voice from the first module to the last, so learners never hear the narrator change.

“Employers have a responsibility to provide a safe workplace and protect workers against possible hazards. You'll see that training workers, pre-job planning and taking the right precautions save lives.”
HeartAmerican English · female
MichaelAmerican English · male
EmmaBritish English · female
GeorgeBritish English · male

Stock synthetic voices from the open Kokoro-82M model. We never clone a voice or make one sound like a real, named person.

What lands in your inbox

A track you don't have to touch.

Every script comes back as a finished, checked track and a short delivery note. Nothing to level, trim or re-record.

WAVmodule-03.wav48 kHz · 24-bit · mono
MP3module-03.mp3192 kbps · 48 kHz
NOTEmodule-03_delivery-notereadings · questions
Every word, in order
Nothing dropped, added or reworded. If your script has a likely typo, we read it as written and ask.
Read like a narrator
Numbers, units, dates, acronyms, phone numbers and web addresses each get one deliberate reading, the same every time.
Every reading listed
The delivery note lists each choice we made. Want “N I S T” instead of “nist”? Say so and we re-render it free.
Mastered to fit the edit
−16 LUFS integrated, true peak at or below −1.5 dBTP, no clipping, half a second of clean silence at head and tail, paragraph pauses you can cut on.
Checked before it ships
Each track is transcribed back to text by three speech-recognition models and compared word for word with your script. Any difference is fixed or explained in the note.
Your glossary, kept
Send product names and jargon once. We keep your pronunciation list from month to month, so module twelve sounds like module one.

How it works

Four steps, about two days.

  1. 01

    Subscribe

    $199 for the month, by card through Stripe. No setup fee, no contract.

  2. 02

    Send scripts

    Email Word, Google Docs or plain text to your Tovrin address, with your chosen voice and any glossary of names and terms.

  3. 03

    We render and check

    Our AI pipeline narrates, masters and checks every track against your script, and writes the delivery note.

  4. 04

    Drop it in

    Files arrive within 2 business days. Ask for any reading to change and we re-render it free.

One plan

$199/month

Up to about 60 finished minutes of narration a month.

  • Finished WAV + MP3 for every script
  • Delivered within 2 business days
  • Every reading listed; any reading re-rendered free
  • Your pronunciation glossary kept month to month
  • Monthly. No contract, no setup fee. Cancel any time.
  • If we can't deliver a track as promised, we refund that month the same day.
Start for $199

Secure checkout by Stripe. After you subscribe, we email you from hello@gettovrin.com to collect your first scripts and voice choice. Currently for businesses in the United States.

How far do 60 minutes go?

Our samples run at about 150 spoken words a minute. Paste a script or type a word count.

Or paste a script to count it
One script6.0 min
Scripts a monthabout 10

For scale: our OSHA sample is 912 words and 6.2 finished minutes, so 60 minutes is about nine modules like it, or fifteen 4-minute explainers.

Built for

  • Bespoke e-learning studios
  • LMS partners who build courses for clients
  • Explainer and animation studios
  • Training teams turning SOPs and policies into video

Not for

  • Cloning a voice, or sounding like a named person
  • Political ads
  • Robocalls or any outbound phone calls
  • Scripts you don't have the rights to use

Questions

Straight answers.

Is this a human narrator?

No. Every Tovrin track is AI-synthesised narration in a stock synthetic voice. Narrating, mastering and checking are done by our AI pipeline, and we never present a track as a human recording.

Do I need to tell learners the voice is AI?

That depends on where you publish and on the law where your audience is. Some platforms and some laws require AI-generated audio to be disclosed. It's your step to take when you publish, and every delivery note reminds you to check.

What counts as a minute?

Finished minutes of delivered audio, measured on the file we send you. A 912-word script came out at 6.2 minutes. Re-renders of a reading you asked to change don't count again.

What if I need more than 60 minutes in a month?

Write to us before you send the extra scripts and we'll agree it with you first. We won't bill you for anything you didn't agree to.

How do you handle our product names and jargon?

Send a glossary when you start, or just send scripts: we pick one reading for each term, list it in the delivery note, and add your corrections to your glossary so it's read the same way in every later track.

Who owns the audio?

You do, as far as we can give it: we assign you all our rights in the tracks we make for you. Fully AI-generated audio may not be protected by copyright, so we can't promise copyright in it. You confirm you have the rights to every script you send.

Can you match our current narrator's voice?

No. We use stock voices only and never clone or imitate a real person's voice. Pick one of our voices and we'll keep it consistent across your whole course.

How do I cancel?

It's monthly with no contract. Email hello@gettovrin.com and we cancel before your next renewal.

Where do the samples come from?

They're made from public-domain US government scripts (OSHA, NASA and NIST) through exactly the same process a paying customer gets. None of those agencies endorses Tovrin.

Talk to us first

Tell us what you're narrating.

Not ready to subscribe? Tell us about your scripts and we'll reply by email with what a month of Tovrin would look like for you.

hello@gettovrin.com

We use what you send only to reply to you. See our privacy notice.