HeyGen
HeyGenAvatar video generation with translation and lip sync
Real-time AI video humans you can talk to, plus an API for building face-to-face video agents into products.
Tavus builds conversational video: AI replicas that see, listen and answer on a live video call. Its own model stack covers full face rendering (Phoenix-4), visual perception (Raven-1) and turn-taking (Sparrow-1), joined to text-to-speech, speech recognition and a language model in one pipeline. Developers create a custom replica of a real person from about two minutes of video, give it a knowledge base of PDFs, slides, images or web pages, set objectives and guardrails, and call tools or their own LLM. A consumer side, PALs, offers AI companions for chat, voice and video calls. It suits product teams building interview, tutoring, sales or support agents with a human face, and differs from avatar video makers by working live and two-way rather than rendering a fixed clip.
Avatar video generation with translation and lip sync
Enterprise avatar video platform for training and internal communication
Talking-avatar video platform that turns a script or a single photo into a presenter video, with translation and interactive agents.
Google's filmmaking tool built on the Veo video models