UNITH launches DEVA-1, a generational leap in Digital Human AI expressiveness
On 3 August 2026, Unith Ltd (ASX: UNT | FWB: CM3) announced the alpha release of DEVA-1, its new state-of-the-art (SoTA) generative visual synthesis AI model designed to significantly enhance the realism and emotional expressiveness of Digital Human avatars.
The release represents a complete overhaul of the synthesiser component within UNITH’s Digital Human platform, engineered to make avatars convey complex emotional messages through facial expressions. UNITH is positioning the model to capture increased share of a Digital Humans market segment currently valued at US$7–10 billion.
The company has set out a defined commercial roadmap, targeting full production rollout by the end of calendar year 2026.
When big ASX news breaks, our subscribers know first
What DEVA-1 delivers — the technology behind the leap
Described as the most expressive model developed by UNITH to date, DEVA-1 follows an extensive review of the latest research in the field. The model has been built around three core capabilities:
-
Enhanced Realism — generation driven by single-image or video inputs.
-
Emotional Responsiveness — improved 3D-subject specific rendering that responds to the tone and characteristics of speech.
-
Optimised Performance — enhanced facial and head movement with latency optimised for real-time conversational applications.
The DEVA-1 optimised performance capability builds on a trajectory of latency improvements already demonstrated in the company’s technology stack, with the Streaming Avatars launch in March 2026 having achieved an 80% latency reduction and sub-one-second response times across real-time conversational applications.
For investors, the significance lies in where this capability is aimed. UNITH has engineered DEVA-1 to address demand from current and prospective clients operating across HR automation, enterprise LMS, hospitality, and B2B sales verticals, segments where lifelike, emotionally responsive avatars can influence engagement outcomes.
| Feature | What It Does | Why It Matters for Clients |
|---|---|---|
| Enhanced Realism | Generates avatars from single-image or video inputs | Lowers the barrier to creating lifelike Digital Humans at scale |
| Emotional Responsiveness | 3D-subject specific rendering that responds to speech tone and characteristics | Enables avatars to convey complex emotional messages, deepening engagement |
| Optimised Performance | Enhanced facial and head movement with latency optimised for real-time use | Supports high-quality conversational applications across enterprise verticals |
Digital Humans explained — why this market matters
A Digital Human, sometimes called an AI avatar, is a lifelike, multilingual, interactive conversational agent. These agents are used to enhance customer engagement, education and entertainment, interacting with users in a realistic manner rather than through static text or menus.
Emotional expressiveness is the differentiator that drives adoption. An avatar that responds to the emotion, tone and cadence of speech feels closer to a human interaction, which is precisely the capability UNITH is targeting with DEVA-1.
The addressable segment is already substantial. UNITH cites a market currently valued at US$7–10 billion and expected to experience significant growth over the coming years. Management has framed DEVA-1 as its lever to capture an increased share of this expanding segment and to pursue category leadership.
Early customer validation signals commercial demand
Alpha testing was conducted with eight selected customers, and initial feedback confirmed that DEVA-1 is viewed as a significant technological leap over the current solution. According to UNITH, all alpha-program partners regarded the model’s realism and ease of deployment as a genuine “generational leap.”
Several key partners have indicated immediate interest in migrating their existing deployments to the DEVA-1 architecture, an early signal of commercial pull.
UNITH intends to pursue a dual-track approach. The company will continue to maintain existing technology deployments where they remain operationally efficient, while simultaneously scaling DEVA-1. The stated aim is a future where customers can select the synthesis model that best aligns with their specific business needs and technical requirements.
Rakan Sleiman, General Manager, Digital Humans
“Our new DEVA-1 speech-driven generative AI model is much more than just a marginal improvement. It represents a significant leap in our technology stack. By enabling our Digital Humans’ facial expressions and visual delivery to respond dynamically to the emotion, tone and cadence of speech, we are directly addressing client demand for more realistic Digital Human avatars. This alpha release is an important step in our mission to become a category leader as an already large Digital Human market segment continues to expand at pace.”
The roadmap to full production by end of 2026
UNITH has set out a defined path from alpha to commercial release:
-
Now – September 2026: Alpha phase continues, with focus on refining DEVA-1’s consistency and adding enterprise-grade features such as branding and subtitle integration.
-
September 2026: Beta release, enabling production-ready access for paying clients.
-
End of calendar year 2026: Full production rollout is currently expected.
Beyond these milestones, UNITH has flagged future enhancement priorities including operational efficiency, scalability, background customisation, and upper body rendering.
For investors, the near-term catalyst is clear. The scheduled beta release is expected to unlock production use for paying clients, providing a defined point at which the technology moves from testing into a revenue-generating context.
Why DEVA-1 strengthens the UNITH investment case
The alpha release brings together a SoTA product, early validation from eight alpha-program partners, a defined roadmap to commercialisation, and exposure to a large, growing market segment valued at US$7–10 billion.
DEVA-1 also aligns with UNITH’s broader strategy of monetising its proprietary AI and digital human capabilities. The company operates a growing B2C subscription division that leverages thousands of UNITH-created digital humans to generate recurring revenue through subscription models.
With no financial figures disclosed alongside the release, the investment case rests on execution against the stated roadmap, particularly the transition to paying-client access at the beta stage and the targeted full production rollout by the end of 2026.
Don’t Miss the Next ASX Tech Breakout
Big News Blast delivers FREE breaking ASX tech news directly to your inbox within minutes of release, complete with in-depth analysis. Join 20,000+ subscribers already staying ahead of the market. Click the “Free Alerts” button at StockWire X to get the next market-moving tech announcement the moment it drops.