Tavus

Tavus's VEED case study describes the deployment of personal avatars using Tavus APIs, allowing VEED to bring the feature to market faster than building it internally. No adoption count or numerical time saving is inferred from VEED's total user base.

Tavus reports more than 2 billion interactions with its video agents and over 100,000 developers using its conversational-video APIs and platform.

Tavus reports more than 2 billion interactions with its video agents and over 100,000 developers using its conversational-video APIs and platform.

Tavus Griffin unifies real-time video-to-video perception and generation in a Human Interaction Model that sees, hears, speaks and moves simultaneously; its Phoenix 4.5 rendering model runs at 134ms latency, Raven processes voice, video and screen context, and Griffin achieved a 48% human-identification rate in a live face-to-face study.

Tavus develops proprietary Phoenix rendering, Raven perception and Sparrow dialogue models for low-latency multimodal conversational video agents.

Tavus develops proprietary Phoenix rendering, Raven perception and Sparrow dialogue models for low-latency multimodal conversational video agents.

Tavus reports a November 2025 $40 million Series B led by CRV, with Scale Venture Partners, Sequoia, Y Combinator, HubSpot Ventures and Flex Capital participating.

Tavus supports enterprise and developer deployments across recruiting, healthcare, sales, learning and product experiences with scalable APIs and enterprise plans.

Tavus supports enterprise and developer deployments across recruiting, healthcare, sales, learning and product experiences with scalable APIs and enterprise plans.

Tavus documents structured conversation callbacks for lifecycle events, transcripts and recording delivery, including common event identifiers and storage-provider metadata. This supplies concrete support for an analyst inference of portable event-driven integration and observability skills beyond broad multimodal expertise.

Real-time video, vision, speech, rendering and agent-orchestration skills transfer broadly across multimodal AI, agents and interactive systems.

Real-time video, vision, speech, rendering and agent-orchestration skills transfer broadly across multimodal AI, agents and interactive systems.