Media platform Josh Talks has developed what it calls the world’s first full-duplex conversational AI model in Hindi. The system, capable of listening and speaking simultaneously, enables voice interactions that more closely resemble natural human conversation.
Unlike traditional voice assistants that operate in a turn-based format, where users must finish speaking before the system responds, the Josh Talks model can listen and speak at the same time. This allows it to acknowledge users mid-sentence, respond in real time, and handle interruptions without breaking the flow of dialogue, the company stated.
“Voice AI has learned to recognise speech and generate speech. What it has not yet learned well is how to participate in conversation. This research shows that when models are trained on large-scale natural dialogue, they can begin to learn the rhythm of how people actually speak to each other,” Shobhit Banga, Co-founder of Josh Talks, said in a statement.
“By building this infrastructure using Hindi conversations, we have demonstrated that full-duplex conversational systems can be developed beyond English,” he added.
The model, named Human-1, was trained on 26,000 hours of natural two-person Hindi conversations involving 14,695 unique speakers. Unlike scripted datasets, this collection captured overlapping speech, interruptions, pauses, conversational acknowledgements, and spontaneous reactions, the company noted. Such data is critical for advancing voice AI, as real-world conversations rarely follow rigid patterns.
Josh Talks has open-sourced the Hindi duplex conversational model to encourage further research in conversational voice AI for Indian languages. While the underlying dataset remains proprietary, the release allows researchers and developers to experiment with full-duplex dialogue systems and build upon the approach.
While the current model focuses on Hindi, the methodology is designed to be replicable across languages, it added. India’s linguistic diversity presents both challenges and opportunities, and Josh Talks believes that large-scale conversational datasets can eventually support natural voice interaction across multiple Indian languages.
By focusing on authentic conversational behaviour, Human-1 moves beyond the limitations of traditional voice systems, adapting to the way humans communicate, the media edtech company boasted.
- OpenAI's Codex Helps in Aerodynamic Engineering for Mazda
- Bengaluru Accounts for 34% of India's FDE Talent: Report
- 84% See AI Productivity Gains, but Only 10% Allow Agent Autonomy: TCS Study
- Snowflake Introduces Business Ontology, Cortex AI Gateway to Bring AI Agents Into Enterprise Workflows
- Three Members Behind Claude Code Leave Anthropic Within 10 Days
- Happiest Minds Q2 Revenue Rises 12.5%; AI Revenue Figure Remains Undisclosed

