Buying AI training data from Nigeria? Email [email protected]Request a free sample

Nigerian-accented English (en) · Lagos, recording to order

Nigerian-accented English speech data for accent-robust voice AI.

English as Nigerians speak it, with accents shaped by Yorùbá, Igbo, Hausa, Pidgin and other first languages. Read prompts and natural conversation from ID-checked, consenting speakers, with first language recorded for every clip.

Email [email protected] Request a dataset

Free 10-hour sample first. A person replies within one working day.

A Nigerian woman speaking into her phone at home

Where it is spoken

Nigerian-accented English at a glance.

English is Nigeria’s official language, used in schools, government, business and media, and spoken with accents that follow each speaker’s first language and region.

Code: en (tagged en-NG in metadata)
Spelling: Standard English spelling; Nigerian English words and expressions kept as spoken
Code-switching: Switches into Pidgin, Yorùbá, Igbo or Hausa transcribed and tagged per word on request
Metadata: First language, region, age band, gender, education band (optional), device

Why it is hard for models

What speech models get wrong.

Accent gaps in big models

Speech models trained mostly on US and UK English make more errors on African accents, a gap research groups have documented.

Source: AfriSpeech-200 (Intron Health), Hugging Face

Nigerian vocabulary

Words and expressions such as “flash me”, “go-slow” and local names need to be in the training text, not treated as errors.

Mixed-language turns

Real conversations slide between English and Pidgin or a local language, often within one sentence.

ASR and TTS

Read prompts or conversation? Usually both.

For speech recognition (ASR)

For accent-robust ASR we record a broad mix of first languages and regions on read prompts and conversation, with first language in the metadata so you can measure error rates per accent.

For text-to-speech (TTS)

For a Nigerian English TTS voice we record a few speakers for longer in quiet rooms, with voice-synthesis consent named explicitly.

Speaker diversity

A fair mix of real speakers.

Speakers balanced by first language (Yorùbá, Igbo, Hausa, Pidgin, others) to your quota
Age band and gender balance agreed up front
Every speaker ID-checked so one voice is one person
Service, customer-support, banking, health and everyday topics

Recording spec

How each clip is recorded and delivered.

Audio: WAV 48 kHz, 16-bit, mono per speaker (one track each in two-person sessions)
Tasks: Read prompts (your list or ours), natural two-person conversation, scripted service dialogues
Transcripts: UTF-8, one per clip, speaker turns and non-speech events marked; second-pass checks on every batch
Metadata: Language, age band, gender, region, device, consent ID and consent version per clip
Consent: Recorded in the speaker’s own language; voice synthesis (TTS) named explicitly when it applies
Delivery: Weekly batches over signed links, with a manifest and a QA report

Open data first

Honest advice: what to try before buying.

Start with open benchmarks

AfriSpeech-200 from Intron Health is an open accented-English dataset that includes Nigerian accents, useful for benchmarking. We supply licensed, consented data to your spec when you need more.

Source: AfriSpeech-200 on Hugging Face

Questions

Buyers usually ask.

Can I measure my model’s error rate by accent?

Yes. Every clip carries the speaker’s first language and region, so you can split your test set by accent group and compare.

Do you record domain-specific English, like banking or health?

Yes. Send us your prompt list or topics and we brief speakers; conversation tasks can be set in a domain too.

How do I order Nigerian-accented English speech data?

Email [email protected] or use the request form with the hours, read/conversation split, speaker mix and transcript rules you need. We record a free sample of about 10 hours to your spec first. Pricing is quoted per approved hour after scoping.

How do speakers consent?

Each speaker passes an ID check and records consent in their own language before any paid task. If the data will be used for voice synthesis (TTS), that use is named explicitly. Speakers can withdraw from future deliveries.

Keep exploring

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.

Email us Request form