Tavus
Tavus builds real-time conversational video AI: developer APIs for lifelike AI agents and replicas, and consumer AI companions it calls PALs.
About this tool
Tavus works on what it calls human computing: AI that can see, hear and respond in real time through a lifelike video face and voice. Developers use Tavus APIs and tools to build conversational video experiences in their products, such as AI interviewers, tutors or support agents, and can train custom replicas based on a real person.
Tavus also offers a consumer product, PALs, described as personal AI companions that listen, remember and respond like a partner, expressing emotion and moving naturally. Developer pricing starts with a free tier including a small number of conversational video minutes, with paid tiers adding minutes and custom replica trainings.
What you can do
Conversational video API
Build real-time AI agents that talk face to face with users on video.
Custom replicas
Train digital replicas of real people for personalised video agents.
PAL companions
Consumer AI companions that remember conversations and respond expressively.
Developer tiers
Free testing minutes and paid plans for scaling conversational video.
Who is it for?
- A recruiting platform runs first-round interviews with a video AI agent.
- An edtech startup builds a face-to-face AI tutor for language practice.
- A consumer chats with a PAL companion as a conversational partner.
Before you choose
Creating replicas of real people requires their explicit consent and clear disclosure to viewers.
Emotional companion products raise wellbeing considerations; set expectations for users.
Frequently asked questions
What is a Tavus PAL?
A PAL is Tavus’ consumer AI companion that listens, remembers and responds in real time with expressive video and voice.
Can developers build with Tavus?
Yes. Tavus provides APIs and tools for real-time conversational video agents and custom replica training.
Is there a free Tavus plan?
Developers can start with a free tier that includes 25 minutes of conversational video, with paid tiers for more usage.
Check the official website for current features, pricing, and terms.
Similar tools
D-ID
VideoD-ID turns photos, scripts and audio into talking-avatar videos and real-time interactive agents, available in its studio and through an API.
Google AI Studio
CodingGoogle AI Studio is Google’s web workspace for testing Gemini models, getting a Gemini API key, building prototypes and moving them into production.
Praktika
EducationPraktika is a language learning app with AI tutors for speaking practice, personalised study plans and real-time corrections.
Elai.io
VideoElai.io is an AI video generator for learning and development that creates presenter videos with avatars, voices and languages for training content.
Wan
VideoWan is a family of video generation models from Alibaba’s Wan team, with open-source releases on GitHub and a hosted creative platform now on Wan 3.0.
WThumb
VideoWThumb creates YouTube and Shorts thumbnail designs from a link, image or written brief, with optional face references and revisions requested in plain language.