Intron Voice AI has launched Sahara v2.5, an upgraded version of its voice artificial intelligence platform designed to improve support for Africa’s multilingual communication patterns.
The latest release supports 63 languages and introduces bilingual conversations across 12 African language pairs, enabling voice AI systems to better handle the way users naturally switch between languages during conversations.
Sahara v2.5 also features what Intron describes as the world’s first trilingual African automatic speech recognition (ASR) model supporting Kinyarwanda, English and French. The model is designed to recognise multilingual speech across the three languages, reflecting the linguistic environments in which many African users communicate.
A central feature of the release is its focus on code-switching. Rather than requiring users to remain within a single language, Sahara v2.5 is designed to recognise and reproduce conversations where speakers move naturally between languages.
Intron has integrated these capabilities across both automatic speech recognition and text-to-speech. This allows the platform to not only understand multilingual speech but also generate voice responses that better reflect users’ linguistic preferences and communication patterns.
The development addresses a longstanding challenge in voice AI, where systems trained primarily around dominant global languages can struggle with African languages, accents and multilingual conversations.
By expanding language coverage and introducing more sophisticated code-switching capabilities, Intron is seeking to make voice AI more relevant to Africa’s diverse linguistic environment.
The launch also reflects the broader growth of African-focused AI infrastructure and language technologies, as developers increasingly build models that account for the continent’s unique languages, datasets and user behaviours.
With Sahara v2.5, Intron is positioning multilingual speech technology as an important component of the next generation of AI services, potentially supporting applications across customer service, education, financial services, healthcare, government and other sectors where voice interaction can improve digital accessibility.
The expansion of African language capabilities could also help reduce language barriers in digital services and enable more users to interact with AI systems using the languages they naturally use in everyday communication.
