Tavus Overview
Tavus is a San Francisco-based AI research lab pioneering Human Computing — teaching machines the art of being human.
Mission
"Automation has scaled efficiency but stripped away empathy, nuance, and presence from digital interactions."
Tavus exists to close the gap between humans and machines by creating AI that can see, hear, understand, and respond with emotional intelligence in real time. Not chatbots with faces — authentic, face-to-face digital presence.
What is Human Computing?
Human Computing is a paradigm shift: computing that adapts to humans, not the other way around.
Core principles:
- Human UI: Interact with AI as naturally as talking to another person — no commands, no learning curve
- Presence over automation: AI that feels like someone, not something
- Emotional intelligence: Reading tone, expressions, context — not just words
The Conversational Video Interface (CVI)
CVI is Tavus's flagship product — an API-first platform for real-time, face-to-face AI conversations.
What makes CVI different from chatbots/avatars:
- Real-time interactive conversation (not pre-rendered video)
- ~600ms latency utterance-to-utterance
- Reads facial expressions, interprets tone, adapts in real-time
- Full orchestration: function calling, RAG, memories
- White-labeled, embeddable, enterprise-ready
CVI Components:
| Component |
What it does |
| Replica |
The visual avatar — your AI's face and appearance |
| Persona |
Behavior, personality, LLM config, system prompt |
| Conversation |
A live WebRTC session connecting replica + persona |
The Model Stack
Tavus builds proprietary models that work together to create human presence:
Phoenix-4 (Rendering)
Gaussian-diffusion model for photorealistic face rendering. Synthesizes high-fidelity facial behavior with:
- Micro-expressions and subtle movements
- Full-face animation (not just lips)
- Real-time emotional response
- Identity preservation
Raven-1 (Perception)
Multimodal perception model that lets AI "see":
- Reads facial expressions and body language
- Detects emotions and intent
- Analyzes environment and screen content
- Contextual awareness
Sparrow-1 (Turn-Taking)
Transformer-based dialogue model for natural conversation flow:
- Knows when to listen, pause, or speak
- ~600ms response latency
- Handles interruptions naturally
- Multilingual support
Products
Conversational Video Interface (CVI)
API-first platform for developers to embed real-time AI video conversations.
- Full pipeline: perception → STT → LLM → TTS → rendering
- Customizable layers (bring your own LLM/TTS)
- Knowledge base (RAG) and memories
- Function calling for external integrations
Video Generation API
Async video generation from scripts or audio.
- Personalized videos at scale
- Custom backgrounds and watermarks
- Transparent background support
Replica API
Create digital twins from 2 minutes of training video.
- Studio-grade fidelity
- Stock replicas available
- Identity preservation
PALs (Personal AI Lifeforms)
Consumer-facing AI companions that remember, evolve, and connect.
- Text, call, or video chat
- Persistent memory
- Proactive check-ins
Use Cases
- Sales & Recruiting: AI SDRs, interviewers, qualification flows
- Education: Tutors, trainers, onboarding
- Healthcare: Patient companions, training simulations
- Customer Support: 24/7 face-to-face assistance
- Personal: Companions, coaches, productivity assistants
Key Stats
- 2B+ interactions powered
- ~600ms utterance-to-utterance latency
- 30+ languages supported
- SOC 2, GDPR, HIPAA compliant (enterprise)
Links
Platform & Docs
Resources
Community
Getting Started
Company
- Founded: 2021 by Hassaan Raza & Quinn Favret
- HQ: San Francisco
- Backed by: Sequoia, Scale Venture Partners, Y Combinator, HubSpot
- Category: Human Computing / AI Research Lab
Pricing Tiers
| Tier |
Minutes |
Replicas |
Concurrency |
| Free |
25 |
Stock only |
1 |
| Starter ($59/mo) |
100 |
1 custom |
3 |
| Growth ($397/mo) |
1,250 |
3 custom |
15 |
| Enterprise |
Custom |
Custom |
Custom + SLAs |
Ethics & Trust
Tavus is built on:
- Informed consent: Every likeness used with permission
- Transparent systems: No hidden levers
- Full disclosure: You know how the magic works
- Bias reviews: Active monitoring and advisory oversight
1---2name: tavus-overview3description: Overview of Tavus, the AI research lab pioneering human computing. Use when you need context about what Tavus is, their mission, core concepts like CVI and Human Computing, the model stack (Phoenix, Raven, Sparrow), or links to docs/platform/resources.4---56# Tavus Overview78Tavus is a San Francisco-based AI research lab pioneering **Human Computing** — teaching machines the art of being human.910## Mission1112> "Automation has scaled efficiency but stripped away empathy, nuance, and presence from digital interactions."1314Tavus exists to close the gap between humans and machines by creating AI that can see, hear, understand, and respond with emotional intelligence in real time. Not chatbots with faces — authentic, face-to-face digital presence.1516## What is Human Computing?1718Human Computing is a paradigm shift: computing that adapts to humans, not the other way around.1920**Core principles:**21- **Human UI**: Interact with AI as naturally as talking to another person — no commands, no learning curve22- **Presence over automation**: AI that feels like *someone*, not *something*23- **Emotional intelligence**: Reading tone, expressions, context — not just words2425## The Conversational Video Interface (CVI)2627CVI is Tavus's flagship product — an API-first platform for real-time, face-to-face AI conversations.2829**What makes CVI different from chatbots/avatars:**30- Real-time interactive conversation (not pre-rendered video)31- ~600ms latency utterance-to-utterance32- Reads facial expressions, interprets tone, adapts in real-time33- Full orchestration: function calling, RAG, memories34- White-labeled, embeddable, enterprise-ready3536**CVI Components:**37| Component | What it does |38|-----------|--------------|39| **Replica** | The visual avatar — your AI's face and appearance |40| **Persona** | Behavior, personality, LLM config, system prompt |41| **Conversation** | A live WebRTC session connecting replica + persona |4243## The Model Stack4445Tavus builds proprietary models that work together to create human presence:4647### Phoenix-4 (Rendering)48Gaussian-diffusion model for photorealistic face rendering. Synthesizes high-fidelity facial behavior with:49- Micro-expressions and subtle movements50- Full-face animation (not just lips)51- Real-time emotional response52- Identity preservation5354### Raven-1 (Perception)55Multimodal perception model that lets AI "see":56- Reads facial expressions and body language57- Detects emotions and intent58- Analyzes environment and screen content59- Contextual awareness6061### Sparrow-1 (Turn-Taking)62Transformer-based dialogue model for natural conversation flow:63- Knows when to listen, pause, or speak64- ~600ms response latency65- Handles interruptions naturally66- Multilingual support6768## Products6970### Conversational Video Interface (CVI)71API-first platform for developers to embed real-time AI video conversations.72- Full pipeline: perception → STT → LLM → TTS → rendering73- Customizable layers (bring your own LLM/TTS)74- Knowledge base (RAG) and memories75- Function calling for external integrations7677### Video Generation API78Async video generation from scripts or audio.79- Personalized videos at scale80- Custom backgrounds and watermarks81- Transparent background support8283### Replica API84Create digital twins from 2 minutes of training video.85- Studio-grade fidelity86- Stock replicas available87- Identity preservation8889### PALs (Personal AI Lifeforms)90Consumer-facing AI companions that remember, evolve, and connect.91- Text, call, or video chat92- Persistent memory93- Proactive check-ins9495## Use Cases9697- **Sales & Recruiting**: AI SDRs, interviewers, qualification flows98- **Education**: Tutors, trainers, onboarding99- **Healthcare**: Patient companions, training simulations100- **Customer Support**: 24/7 face-to-face assistance101- **Personal**: Companions, coaches, productivity assistants102103## Key Stats104105- 2B+ interactions powered106- ~600ms utterance-to-utterance latency107- 30+ languages supported108- SOC 2, GDPR, HIPAA compliant (enterprise)109110## Links111112### Platform & Docs113- **Homepage**: https://www.tavus.io114- **Developer Portal**: https://platform.tavus.io115- **Documentation**: https://docs.tavus.io116- **API Reference**: https://docs.tavus.io/api-reference/overview117- **CVI Overview**: https://docs.tavus.io/sections/conversational-video-interface/overview-cvi118119### Resources120- **Research**: https://www.tavus.io/research121- **Blog**: https://www.tavus.io/blog122- **Example Projects**: https://github.com/Tavus-Engineering/tavus-examples123- **Status**: https://status.tavus.io124125### Community126- **Discord**: https://discord.gg/5Y9Er6WNN5127- **GitHub**: https://github.com/Tavus-Engineering128129### Getting Started130- **Sign Up**: https://platform.tavus.io/auth/sign-up?is_developer=true131- **Free Tier**: 25 conversational minutes + stock replicas132133## Company134135- **Founded**: 2021 by Hassaan Raza & Quinn Favret136- **HQ**: San Francisco137- **Backed by**: Sequoia, Scale Venture Partners, Y Combinator, HubSpot138- **Category**: Human Computing / AI Research Lab139140## Pricing Tiers141142| Tier | Minutes | Replicas | Concurrency |143|------|---------|----------|-------------|144| Free | 25 | Stock only | 1 |145| Starter ($59/mo) | 100 | 1 custom | 3 |146| Growth ($397/mo) | 1,250 | 3 custom | 15 |147| Enterprise | Custom | Custom | Custom + SLAs |148149## Ethics & Trust150151Tavus is built on:152- **Informed consent**: Every likeness used with permission153- **Transparent systems**: No hidden levers154- **Full disclosure**: You know how the magic works155- **Bias reviews**: Active monitoring and advisory oversight
tavus-engineering