01
Fund your account
Trainers top up their wallet with tokens, which pay for the AI-assisted tools used to generate, transcribe, and analyze prompts during training.
About Dialect Library
Building voice AI with the people who know each dialect best.
Dialect Library is a crowdsourced voice and dialect data platform. Trainers record prompts, translate word tasks, and help create speech datasets that reflect how people actually speak in their communities.
How it works
From funding your account to getting paid, referring others, or cashing out — here is the full loop.
02
Translate and record
You receive English words or prompts, translate them into your dialect, and record voice samples in languages like Igbo, Hausa, Yoruba, Tiv, and others.
03
Get scored and paid
Submissions are quality-scored and paid out from the pool funded by data subscribers, covering your token spend and putting you in profit.
04
Refer, resell, or cash out
Earn admin-configured referral bonuses from confirmed token funding and eligible scored training payouts from people you invite.
Payout scoring, the subscriber-funded reward pool, and member-to-member token resale are actively being built out. Token funding is live today, and referral bonuses are controlled from the admin dashboard.
Dialect-first data
Dialect Library is built around local speech, not generic language labels. Trainers contribute the words, accents, and phrasing that speech systems usually miss.
Human review
Submissions are designed to be cross-checked by other trainers in the same dialect cluster before they count toward reward calculations.
Ongoing transparency
The platform is under active development. Totals, reward flows, and scoring views will keep evolving as more submissions and settlement data come in.
Why this matters
Speech tools often work best for well-represented accents and languages. Dialect Library focuses on collecting high-signal local examples so future ASR and voice systems can serve more speakers accurately.