Dataset catalogue

Example Speech Dataset Specifications

These are example specifications, not off-the-shelf inventory. Each one shows how a dataset can be scoped; the final language, speaker counts, hours and licensing are defined with you before collection begins.

Example datasets

How a specification is described

Every field below is confirmed per project. Nothing is claimed as already recorded or already licensed.

Example dataset

Thai — Read speech / ASR

Language
Thai
Type
Read speech / ASR
Speakers
To be defined per project
Gender
Balanced or custom quota
Duration
Scoped to requirement
Environment
Quiet indoor / studio
Transcription
Available on request
Metadata
Speaker, session, device
Licensing
Custom, agreed per project
Availability
Custom collection on demand
Request this specification →

Example dataset

Thai — Conversational speech

Language
Thai
Type
Conversational speech
Speakers
Pairs or small groups
Gender
Custom quota
Duration
Scoped to requirement
Environment
Quiet indoor, separate channels possible
Transcription
Speaker-attributed, on request
Metadata
Speaker, topic, turn labels
Licensing
Custom, agreed per project
Availability
Custom collection on demand
Request this specification →

Example dataset

Thai — TTS / single speaker

Language
Thai
Type
TTS / single speaker
Speakers
Single or multi-speaker
Gender
Selected by casting
Duration
Scoped to requirement
Environment
Studio-quality
Transcription
Script-aligned
Metadata
Speaker profile, script IDs
Licensing
AI training rights defined in contract
Availability
Custom collection on demand
Request this specification →

Example dataset

Thai — Wake word & commands

Language
Thai
Type
Wake word & commands
Speakers
Multi-speaker
Gender
Custom quota
Duration
Scoped to requirement
Environment
Near-field, far-field, in-vehicle
Transcription
Prompt-aligned labels
Metadata
Distance, angle, device, noise
Licensing
Custom, agreed per project
Availability
Custom collection on demand
Request this specification →

Looking for something not listed? See custom speech dataset collection or send your specification directly.

Tell us what data you need

Send your dataset requirements — language, speakers, hours, environment, transcription and licensing scope. If the specification is still open, our team can help define it.