Synthesia debuts interactive AI avatars for enterprises and individuals

Synthesia, a digital avatar startup valued at $4 billion, has expanded its platform to include interactive AI avatars that can engage in real-time conversations with users. These avatars are trained on specific content or scripts and are designed to respond to user queries in a conversational manner.

The company’s technology stack for these avatars combines voice-to-text, language models, text-to-voice, and video models. Synthesia’s own video and voice models power the avatars, though customers can integrate alternatives from providers like Cartesia, ElevenLabs, Google, or OpenAI. Enterprises can host the avatars on their preferred cloud infrastructure or opt for Synthesia’s hosting services.

Synthesia’s platform now includes three product types: a video-creation and distribution platform with static avatars, an agentic platform called Sessions for interactive roleplay and surveys, and an API platform for building custom interactive avatars or other products. The interactive avatars are deterministic, meaning they only respond to queries within their trained scope, such as a specific article or script.

The company demonstrated the technology by creating a personal avatar for a TechCrunch journalist, trained on a story about venture fraud. The avatar was built using photos and a two-minute voice recording, with consent obtained for its creation. The process took the Synthesia team a couple of days to complete.

Synthesia’s interactive avatars are part of a broader trend in AI-driven content creation and customer interaction. The company’s valuation reached $4 billion earlier in the year, and it reported crossing $100 million in annual recurring revenue (ARR) last year. Competitors in the digital avatar space include D-ID, HeyGen, and Colossyan.