AI voice licensing: dataset recording plus a licence to use my voice
I don’t just sell recording hours. I record the dataset your model needs and license the use of my voice in your AI (TTS, virtual assistant or customer-service agent), with clear terms for usage, duration and territory.
What’s included
Custom TTS datasets
Scripted reads, single sentences, free conversation and emotions, following your specs and file naming.
Multiple voice profiles
Conversational, neutral, energetic, warm or corporate: different styles from the same voice for a more versatile model.
Assistants & customer service
Voices for virtual assistants, voice bots, IVR and phone systems: clear and friendly.
Three Spanish variants
Neutral Latin American Spanish, Mexican Spanish and the Rioplatense accent of Argentina.
Clean studio audio
48 kHz / 24-bit mono WAV, noise floor below −60 dB, no reverb or processing.
Voice usage licence
The price includes the licence: together we define how the voice is used in your model, for how long and where.
How it works
- 1Specs
You send scripts, technical format, voice profiles and file naming.
- 2Test
I record a sample to validate quality, style and pronunciation before we start.
- 3Dataset recording
I record in my treated booth, in batches, with ongoing quality control.
- 4Delivery
Edited files, named and organised by profile, ready for training.
Frequently asked questions
What audio format do you deliver?
48 kHz / 24-bit mono WAV by default, but I adapt to your project’s specs.
Can you record many hours of audio?
Yes. We plan the volume in batches to keep the voice consistent and the quality even across the whole dataset.
Do you sign consent for AI use of your voice?
Yes. I work with clear consent and licences that define how the voice is used in the model.
Which accents can you record?
Neutral Latin American Spanish, Mexican Spanish and the Rioplatense accent of Argentina and Uruguay.
Let’s talk about your project
Tell me what you need and I’ll reply with a quote and timeline, no strings attached.