Companies

Tavus

tavus.io

Tavus builds human-like AI models, digital twins, and real-time conversational experiences through developer APIs.

HQSan Francisco, California, United States
Employees51-200
Funding$64.2M
Valuation$250.79MM
14 active roles
Profile 6mo agoJobs checked 17h ago
AI / MLFoundation Model ProviderB2B SaaSSeries A$50M-$200M

About

Tavus is an AI research lab building foundational models and real-time AI-human experiences, including digital twins and interactive personal AI companions. It sells developer APIs and tools to developers, founders, product teams, and businesses, differentiating itself through human-like, emotionally intelligent, real-time interaction rather than asynchronous avatar or video generation.

Market

Tavus competes in generative AI video, AI avatar, and real-time conversational-agent infrastructure. It positions itself around “human computing”: API-first, face-to-face AI interactions that combine perception, turn-taking, speech, language models, and expressive real-time rendering, rather than focusing only on pre-rendered or scripted avatar videos. Its differentiation is the CVI stack, WebRTC delivery, low-latency multimodal interaction, customizable RAG and LLM-backed personas, and production-oriented enterprise deployment.

Target Customers

Tavus primarily serves developers, founders, product teams, and enterprise teams that want to embed human-like AI video agents into products, services, or workflows. Its strongest vertical targets include sales, healthcare, learning and development, recruiting, education, and customer-facing applications.

At a Glance

Problem

Tavus addresses the limits of conventional AI interfaces and personalized video: text and voice can be efficient, but they often lack the presence, trust, emotional cues, and human connection of face-to-face interaction. For businesses, creating genuinely personalized video at scale is also expensive and operationally heavy, traditionally requiring recording studios, equipment, editing, and significant employee time. Tavus’s economic proposition is to automate those costs while making individualized engagement scalable.

The clearest initial use case is sales and marketing follow-up: when a prospect submits a form, a company can automatically send a customized video instead of a standard email. The same infrastructure extends to customer support, recruiting, education, healthcare, and training, where more engaging interactions may improve productivity, trust, conversion, or loyalty.

Product / Service

Tavus provides an AI-video infrastructure platform delivered through developer APIs, a no-code portal, and enterprise services. Its models support text-to-video avatars and the Conversational Video Interface, or CVI, which lets developers build real-time conversational AI agents called PALs. The platform combines perception, understanding, voice, and rendering; developers can use a stock persona or create a custom face from a photo or short training video, then deploy it across many conversations.

The real-time stack is designed to make AI agents appear present and emotionally responsive rather than merely animated. Phoenix-4 generates full facial behavior in real time, while Tavus’s broader system handles visual perception, conversational timing, voice, and interaction. The platform supports low-latency 1080p video, multiple languages, custom replicas, and deployment through free, pay-as-you-go, growth, and enterprise plans, with enterprise compliance, integration, and implementation support.

Market

Tavus competes in generative AI video, B2B video infrastructure, conversational AI, and the emerging human-computing or AI-human category. Its nearer-term alternatives include Synthesia, HeyGen, VEED, and Rephrase.ai, although Tavus differentiates itself by emphasizing developer infrastructure and real-time, face-to-face conversational agents rather than only asynchronous avatar or marketing-video generation.

Tavus is an active, venture-backed commercial company rather than an apparent pre-revenue project. Its site says the platform is trusted by enterprises and startups and used by more than 100,000 developers; Y Combinator lists it as an active company, and Tavus announced a $40 million Series B in November 2025. Third-party funding records put cumulative funding at approximately $64.2 million. The reviewed sources do not disclose revenue or a precise paying-customer count, so traction is best characterized by developer adoption, enterprise availability, product launches, and financing rather than reported financial performance.

Founders & Leadership

Hassaan RazaFounder
Co-Founder and CEO
Quinn FavretFounder
Co-Founder and COO
Julia SzatarHead of Marketing
Effie GoenawanHead of Product

Funding History

2021-09
Seed$125K

Y Combinator

2023-03
Seed$6.1M

Sequoia Capital

2024-03
Series A$18M

Scale Venture Partners

2025-11
Series B$40M

CRV

Recent News

2026-04-27partnership
Startups & Grants Program

Tavus promoted its Startups & Grants Program, inviting startups to integrate Tavus directly into their products and add a human-like presence to applications.

2026-02-19product
Tavus Launches Phoenix-4: the First Real-Time Human Rendering Model with Emotional Intelligence

Tavus announced Phoenix-4, a real-time human-rendering model focused on emotional intelligence. The model is designed to work with Tavus’s perception and conversational-timing systems as part of a broader human-computing stack.

2026-02-18product
Phoenix-4: Real-Time Human Rendering with Emotional Intelligence

Tavus published its product announcement for Phoenix-4, describing how it works alongside Sparrow-1 for conversational timing and Raven-1 for perception.

2026-02-10product
Raven-1: Bringing Emotional Intelligence to Artificial Intelligence

Tavus introduced Raven-1, a multimodal perception system intended to capture not only what users say but also how they say it.

2026-01-22product
Build with Tavus

Tavus introduced a platform-oriented offering for building, scaling, customizing, and integrating real-time AI video agents into products and workflows.

2025-11-15funding
Tavus Official Information: $40M Series B

Tavus reported a $40 million Series B led by CRV and approximately $64 million raised to date.

2025-11-13funding
Tavus Raises $40 Million Series B to Advance Emotionally Intelligent AI Humans

Pulse 2 reported that Tavus raised $40 million in Series B funding, led by CRV with participation from Scale Venture Partners and other investors, to advance human-computing and emotionally intelligent AI-human technology.

2025-11-12funding
Tavus Raises $40M to Build the Next Frontier of Intelligence: Human Computing

Tavus announced $40 million in Series B funding led by CRV, with participation from Scale Venture Partners, to build the future of human computing.

2025-09-21
Tavus Research Models Phoenix-3, Raven-0, and Hummingbird-0 Redefine Realism and Perception in AI Systems

A September 2025 article highlighted Tavus’s Phoenix-3, Raven-0, and Hummingbird-0 research models, describing them as reshaping human-AI interaction through improved realism and perception.

2025-08-05partnership
Partner with Tavus

Tavus published a partnership invitation describing collaboration opportunities with consultants, platforms, and agencies to bring Tavus to more users. The page does not identify a specific named partner or integration.

Active Roles

14
San Francisco/Marketing/56d ago
San Francisco/Engineering/56d ago
San Francisco/Product/56d ago
San Francisco/Design/56d ago
San Francisco/Sales/56d ago
San Francisco/Design/56d ago
San Francisco/Sales/57d ago
San Francisco/Engineering/111d ago
San Francisco/Marketing/128d ago
Remote/Engineering/134d ago
Remote/Engineering/136d ago
Remote/Data & Analytics/157d ago
San Francisco/Forward-Deployed Engineer/164d ago
San Francisco/Solutions Engineer/196d ago

Business Model

Tavus monetizes its developer platform through paid monthly subscription tiers, including Starter and Growth plans, supplemented by usage-based API and per-minute charges. Its pricing is designed to scale with access to developer APIs and real-time AI-video capabilities.

Products

Conversational Video Interface (CVI) platform and APIsAI Video Agents and PALs / AI PersonasPhoenix real-time avatar-rendering engine and 1:1 AI Replicas or digital twinsText-to-video AI avatars

Customers

AmazonDeloitteEYMayo ClinicCVSSalesforceAetnaWixMetaOrumDelphiFinal Round AICareFlick

Tech Stack

Multimodal AI models spanning language, audio, and videoWebRTC real-time media, powered by DailySpeech-to-text (STT), large language models (LLMs), and text-to-speech (TTS)Raven visual-perception modelsSparrow conversational-flow and turn-taking modelsPhoenix real-time facial-rendering and avatar engineRetrieval-augmented generation (RAG), custom LLMs, and intelligent function callingAPI-first architecture

Competitors

Synthesia
HeyGen
VEED
Colossyan

Key Investors

Y Combinator, Sequoia Capital, CRV, Scale Venture Partners, Accel, Lightspeed Venture Partners, SV Angel, HubSpot Ventures, Remus Capital, Flex Capital