Unith Unveils DEVA-1 Alpha to Elevate Digital Human Realism

Alpha Release Marks a Leap in Digital Human Technology

Unith Ltd (ASX:UNT) has taken a significant step forward in AI-driven digital human technology with the alpha release of DEVA-1, its latest state-of-the-art generative AI model. Designed to dramatically improve the expressiveness and realism of Digital Human avatars, DEVA-1 supports single-image or video avatar generation and real-time streaming that dynamically responds to speech tone and emotion.

The company describes this release as a "generational leap" in its synthesiser technology, aiming to meet growing client demand for digital avatars capable of conveying complex emotional messages, particularly across sectors like HR automation, enterprise learning management, hospitality, and B2B sales.

Technological Breakthroughs Drive Real-Time Expressiveness

DEVA-1 introduces enhanced 3D rendering tailored to individual subjects, enabling avatars to react visually to nuances in speech tone and cadence. This includes improved facial and head movements with latency optimised for seamless conversational applications. The model’s ability to generate avatars from minimal input, single images or videos, marks a notable advance in user accessibility and deployment speed.

Strong Early Validation from Alpha Partners

Eight selected customers participated in DEVA-1’s alpha testing phase, providing overwhelmingly positive feedback. All partners recognised the model’s realism and ease of integration as a step-change from previous solutions, with several expressing intent to migrate existing deployments to the new architecture. Unith plans to maintain legacy systems for clients where appropriate, while scaling DEVA-1 to meet broader market needs.

Strategic Roadmap Targets Market Expansion

The alpha phase will continue through to September 2026, during which Unith will focus on refining consistency and adding enterprise-grade features like branding and subtitle integration. A beta release is scheduled for the same month, opening production-ready access to paying clients. The company anticipates a full production rollout by the end of 2026, aiming to capture a larger share of the Digital Humans market, currently valued at US$7–10 billion and expected to grow significantly.

Future improvements will prioritise operational efficiency, scalability, background customisation, and upper body rendering, signalling a comprehensive upgrade path beyond the initial facial animation enhancements.

Leadership Sees DEVA-1 as a Market Differentiator

Rakan Sleiman, General Manager of Digital Humans at Unith, highlighted the model’s ability to make avatars respond dynamically to emotional cues in speech as a critical advancement. He emphasised that DEVA-1 is not merely an incremental update but a foundational technology shift positioned to strengthen Unith’s leadership in a rapidly expanding market segment.