The Igbo API | Building Unique Igbo Language Experiences

With over 525 native languages, Nigeria is the 3rd linguistics diverse countries on the platform. It's home to fourth most spoken African language: Igbo. Despite there being more than 35 million speakers, popular language recognition platforms like Google's Assistant, Amazon's Alexa, and Apple's Siri don't offer support for native Igbo speakers to interact with technology with their voice.

More than half the Nigerian population is projected to be connected to the internet by 2027. This is a major sign that there will be more people online that want to be able to community in their primary language.

The Igbo API is the first, open-source, multi-purpose African language API that offers AI features like automatic speech recognition, machine translation, and more allowing developers to build experiences that connect with a broader African population.

Frequently Asked Questions

What features are currently available?

The Igbo API offers four different features: Speech-To-Text transcriptions, Igbo to and from English translations, and an Igbo dictionary via a REST API.

What data was used to train these models?

Our open-source dataset can be found on our Hugging Face. The data was sourced by an open-source community of more than 500 volunteer audio recorders and translators. We were funded in 2021 by the Lacuna Fund that allowed us to build the largest Igbo-English dictionary and a good portion of the data to train our models.

How can I use the Igbo API?

You can get started by signing up for a developer token.
Then you can visit our documentation site to start using the Igbo API.

I speak Igbo, how can I contribute?

We are looking for more native Igbo speakers to record audio and translate sentences. If you would like to contribute to the dataset that will directly improve the quality of the model, you can create an account on Sabbi our data-protected Igbo language data collection platform.