DhivehiGPT Introduces New Dhivehi Transcription and Text-to-Speech Features

A recent report by Voice.mv highlighted a major update to DhivehiGPT, the local artificial intelligence platform created by Javaabu. The update introduces a full transcription system that can convert spoken Dhivehi audio into written text, along with new text-to-speech capabilities that allow the AI to read Dhivehi text aloud.

According to the Voice.mv article, this is the first time Maldivians have access to a reliable Dhivehi speech-to-text tool. While such technology is widely available in other languages, the lack of Dhivehi support has made transcription work slow, expensive and difficult for many professionals. With this new feature, users can upload audio files and receive transcripts that separate individual speakers and can be exported as Word files or subtitle documents. Users can also ask DhivehiGPT to clean, summarise or translate the transcripts directly through its chat interface.

The update also introduces a Dhivehi text-to-speech service. The first available voice is a female voice named Hajja, and Javaabu has stated that more voices are under development. This new feature is expected to be especially useful for people with disabilities, content creators and anyone who needs Dhivehi audio output.

Voice.mv also reported that DhivehiGPT has added several other improvements. These include a new credit system, affordable individual plans, team plans designed for offices, and additional language translation options such as Arabic, Bengali and Hindi. Another long-requested addition, chat history, has also been introduced.

All these features are now available on dhivehigpt.com, and Javaabu believes the update marks an important step in advancing Dhivehi-language technology. The company hopes that easier access, affordable plans and new AI capabilities will encourage wider local adoption, helping the Dhivehi language evolve alongside modern digital tools.