Nigerian speech data, recorded and transcribed to your spec.
Nigerian Pidgin, Yorùbá, Igbo, Hausa and Nigerian-accented English, with the code-switching people really use. Read prompts or natural conversation, from ID-checked speakers who consent on record.
Free 10-hour sample first. A person replies within one working day.
Speech the way Nigerians actually talk.
Read prompts
Your prompt list read aloud, one clip per prompt, for ASR and TTS. Script checks before recording.
Natural conversation
Spontaneous talk between two consenting speakers, each on their own microphone track.
Market and service talk
Buying, selling, bargaining and giving directions: the everyday domain voice agents need.
Accented English
Nigerian English across regions and ages, for accent-robust speech recognition.
Five languages, with code-switching.
- Nigerian Pidgin (pcm)
- Spoken across the country; read and conversational
- Yorùbá (yo)
- South-west; tone-marked transcripts on request
- Igbo (ig)
- South-east; read and conversational
- Hausa (ha)
- North; read and conversational
- Nigerian English (en)
- Accented English, often mixed with the languages above
- Formats
- WAV 48 kHz mono per speaker, JSON metadata, UTF-8 transcript per clip
- Metadata
- Language, speaker age band, gender, region, device, consent ID and version
Start with open datasets. Come to us for what they don’t cover.
Open Nigerian speech corpora are a good place to begin, and you should check each one’s licence for commercial use. We collect new, licensed audio to your spec: your prompts, domains, speaker mix and transcription rules.
- Commercial licence from the start, with a consent record per speaker
- Your domain and vocabulary, not whatever was easy to scrape
- Speaker balance you choose: age, gender, region, language mix
- A person checks every clip before delivery
Buyers usually ask.
Which Nigerian languages do you record?
Nigerian Pidgin, Yorùbá, Igbo, Hausa and Nigerian-accented English, including the code-switching between them. Other Nigerian languages can be scoped on request.
Do you provide transcripts?
Yes. Every clip can ship with a verbatim UTF-8 transcript in the language spoken. Tone marks for Yorùbá and translation layers can be agreed per order.
Is the speech data licensed for commercial AI training?
Yes. Each speaker records a consent statement that their recordings can be licensed to companies that train AI, in Nigeria and abroad. The consent ID and version ship in the manifest for every clip.
Can you record our own prompt list?
Yes. Send the prompts and the speaker mix you need. We check the script, record and review every clip, and send a free sample first.
How do I buy Nigerian speech data?
Email [email protected] or use the request form with the languages, hours, task type and format. A person replies within one working day. Pricing is quoted after scoping.
More from TrustAm Data.
- African language dataLow-resource languages, consented
- Dataset libraryBrowse by trade, language and modality
- Custom data collectionPilot first, then scale
- For AI labsLicence, consent, QA, NDA
- Trust and consentPeople, places, minutes
- Matching datasetsCollect-to-order listings
Tell us what you need. We send a free sample first.
Email [email protected] with the data type, hours, languages or trades, labels and format. Pricing is quoted per approved hour after scoping.