Conversational and scripted speech
Natural dialogue, read speech, wake words and command phrases recorded to defined scripts and audio specifications.
Speech & Voice AI
Speech data for recognition, synthesis and voice assistants
Voice models need variety that is hard to find in public data: accents, emotion, natural conversation and clean, consistent recordings. One6 combines crowd-sourced speech, professional voice talent and audio annotation, recorded and labeled to your technical specification.
What One6 provides
Professional voice talent
Voice actors for narration, expressive and emotional speech, character voices and high-quality TTS data.
Accent and dialect coverage
Speakers recruited across accents, ages and regions to reduce bias in voice models.
Transcription and labeling
Transcription, speaker labeling, timestamps and sound-event tagging for speech and audio models.
Services involved
- Speech & Voice DataConversational and scripted speech, accents, emotion and professional voice talent.
- Data Annotation & LabelingImage, video, text and audio labeling with written guidelines and layered quality review.
- Multilingual AI DataNative-speaker collection, transcription, localization and evaluation across languages.
How a pilot runs
- 01
Specify
Sample rate, format, channels, noise conditions, metadata and speaker profile are written into the project spec.
- 02
Agree consent
Consent and usage terms for every recording are agreed with you and put in place with contributors before collection.
- 03
Pilot
A first set of recordings, checked against the spec and reviewed with you.
- 04
Scale
Recording and annotation in regular batches with ongoing QA.
Frequently asked questions
What audio formats and quality can you deliver?
Sample rate, format, channel setup, noise conditions and metadata are defined in the project spec, and recordings are checked against it before delivery.
Do speakers consent to their voice being used for AI?
Consent and usage terms for every recording are agreed with you during scoping and put in place with contributors before collection begins.
Can you collect speech in lower-resource languages?
Collecting speech for lower-resource languages is one of the projects we support. Tell us the languages and dialects you need and we'll confirm availability during scoping.
Related solutions
Have a data project in mind?
Tell us what your model needs. We'll come back with questions, a proposed approach and a pilot plan.