Tavus Blog · May 2026
Launches Image-to-Replica, letting Tavus train a Phoenix-4 AI human from a single image and expand the set of possible faces that can become real-time conversational AI humans.
Public company, workplace, funding, and market signals
Updated Jul 29, 2026
Tavus is a 2020-founded San Francisco AI research lab building real-time conversational video AI humans (PALs/CVI) for developers and enterprises, with a public Series B and customers across sales, healthcare, recruiting, and education.
Primary product
Conversational Video Interface (CVI) and PALs — real-time AI humans for face-to-face conversations.
Founded
2020
Headquarters
San Francisco, California, United States
Team size
70-80
Industry
Software Development
Sub-industry
Generative AI / conversational video AI / AI humans
Offices
16 jobs at Tavus
Business model
Stage
Series B
Total raised
$64M
Latest round
Series B · Nov 2025
Latest amount
$40M
Mar 2023 · Sequoia Capital
Leadership
Co-Founder & CEO
Co-Founder & COO
Head of Research
Head of Engineering
Head of Sales
Mission-driven and research-heavy, with an emphasis on ownership, shipping fast, and working alongside AI research leaders on human-computing systems.
Compensation
Public careers materials say compensation includes salary and company equity; no public salary bands were found.
Benefits
Pricing
Free, Starter, Growth, and Enterprise tiers with included conversation minutes/faces; enterprise and managed deployments also appear to use custom/volume pricing.
Differentiators
Technology
Customers
Competitors
Estimated revenue
$1M-$10M
Estimated monthly visits
67.4K
Traffic estimate as of Jun 2026
Tavus Blog · May 2026
Launches Image-to-Replica, letting Tavus train a Phoenix-4 AI human from a single image and expand the set of possible faces that can become real-time conversational AI humans.
BusinessWire · Feb 2026
Introduces Phoenix-4, a real-time behavior generation engine for emotionally responsive, context-aware human presence in live conversation.
BusinessWire · Feb 2026
Launches Raven-1 GA, a multimodal perception layer that fuses audio and visual signals to understand emotion, intent, and context in real time.
BusinessWire · Jan 2026
Introduces Sparrow-1, Tavus's conversational-flow control model for low-latency turn-taking and human-like timing in live voice/video interactions.
BusinessWire · Nov 2025
Announces a $40M Series B led by CRV, with participation from Scale Venture Partners, Sequoia Capital, Y Combinator, HubSpot Ventures, and Flex Capital.