Pricing built to scale
Start building conversational video experiences for free,
with flexible pricing as you grow
with flexible pricing as you grow
Free
$0
For developers that want to get their feet wet
Includes:
White-labeled APIs
25 free minutes of AI conversational video
5 free minutes of AI video generation
Access to 25 stock replicas
Support for 30+ languages
Watermark free
Custom Replicas
Starter
Most popular
$59
/mo + pay as you go usage
For builders starting to craft human-AI experiences
Everything in Free, plus:
3 custom replica trainings
100 minutes of AI conversational video
10 minutes of AI video generation
Pay-as-you-go overage with no limits
Up to 3 concurrent streams
Growth
$397
/mo + pay as you go usage
For teams building and productionizing AI conversations
Everything in Starter, plus:
7 custom replica trainings
500 minutes of AI conversational video
100 minutes of AI video generation
Access to 100+ stock replicas
Conversation recordings
Up to 15 concurrent streams
Enterprise
Custom
For scaling businesses and advanced use cases
Everything in Growth, plus:
100% white-labeled experience
Scaling discounts
Custom concurrency limits
Top-tier support
Enterprise-grade security and compliance
Guaranteed SLAs for speed and compute
Compare Plans
Free
Starter
Growth
Enterprise
General Features
API access
Full access to our suite of developer APIs, as well as no-code portal
Watermark-free
Generate videos without a Tavus watermark
White-labeled APIs
Full white-label experience (except for the Replica consent process)
Quickstart toolkit
Use out of the box examples and prebuilt components to accelerate development
Guaranteed uptime and SLAs
Enterprise-grade uptime & performance SLAs
Replicas
Stock Replica library access
Access a diverse library of 100+ high-quality stock replicas for quick testing or usage
25 Replicas
25 Replicas
100+ Replicas
100+ Replicas (including custom)
Replica training API
Train Replicas via APIs. No human in the loop workflow needed
Included custom Replica trainings
Generate hyperrealistic AI replicas of real-life humans from just a two minute video (includes custom voice model)
3 Replica trainings per month
7 Replica trainings per month
Pay-as-you-go Replica training overage
One-time flat fee per additional custom replica created beyond your monthly quota
$65/Replica
$40/Replica
Scaling discounts
White-labeled Replica consent
100% whitelabeled experience, replace Tavus branding with your own company name in the replica consent script workflow
30+ languages
Create seamless AI experiences in 30+ languages
Alpha channel video
Human cutout with a transparent background
Conversational Video Interface (CVI)
End-to-end conversational video pipeline
Out of the box optimized AI pipeline with best-in-class vision, turn-taking, and rendering models, including TTS, LLM, STT, ASR, and WebRTC
Pipeline modularity
Easily customize the CVI pipeline by bring your own models
Included CVI minutes
Allocated minutes for real-time AI-driven conversations, with additional usage billed separately. Each conversation incurs a 30-second minimum charge
25 mins per month
100 mins per month
500 mins per month
Scaling discounts
Pay-as-you-go CVI overage
Pay per additional minute of streamed AI conversations, rounded to the nearest 6 seconds
$0.37/min
$0.32/min
Scaling discounts
Maximum concurrent streams
The maximum number of concurrent CVI streams that can run at the same time
1
3
15
Custom
Maximum conversation duration
The maximum length of a conversation
5 mins
20 mins
60 mins
Custom
High resolution video
Delivers the highest possible fidelity via WebRTC with ultra-low latency and seamless responsiveness
1080p
1080p
1080p
1080p
Full-face generation
Every pixel of the face is generated, ensuring maximum naturalness
Vision
Dynamic low latency
State-of-the-art turn-taking model, which dynamically adapts to speech patterns to provide <500ms utterance to utterance
SOTA contextual understanding
Context aware turn taking to power truly natural conversations
Highest fidelity audio
Crisp and clean audio fidelity
24kHz
24kHz
24kHz
24kHz
Bring your own voice engine
Easily bring your own TTS engine, such as ElevenLabs, Cartesia, or PlayHT
Function calling
Tools for the LLM to perform tasks beyond just text generation. Useful to integrate external APIs or services into the LLM
Bring your own LLM
Easily integrate your own LLM, or modify the LLM used by the Tavus system
Conversation transcripts
Automatically and instantly save a text transcription of every conversation
Conversation recordings
Capture and store video recordings of conversations
$0.03/min
Scaling discounts
Professionally optimized CVI Replicas
Finetuned Replicas specifically optimized for the most natural CVI experience
Video Generation
Included video generation minutes
Allocated video generation minutes per month, with additional usage billed separately. Each conversation incurs a 30-second minimum charge
5 mins per month
10 mins per month
100 mins per month
Scaling discounts
Pay-as-you-go video generation overage
Pay per additional minute of generated video, rounded to the nearest 6-second increment
$1.10/min
$0.90/min
Scaling discounts
Dynamically captured website backgrounds
Generate screenshare-style backgrounds dynamically by just by providing a URL
Content moderation
Safeguards to prevent the creation of harmful content
Enterprise-grade
Priority compute
Power faster AI experiences with high-priority GPU access
Support and Services
Email Support
Responsive human support with resolutions, and creative solutions
Best-effort response times
Standard support
Priority support
Dedicated priority support
Dedicated technical support and private Slack channel
Hands on, use-case tailored support. Includes Private Slack channel support
SOC 2 and HIPAA compliance
Meet enterprise compliance requirements with access to SOC2, HIPPA, and compliance reports
Custom trained models
We tune models tailored to your use case for optimized performance and accuracy
Bespoke development and integration
Custom demos and solutions built with you, from development to deployment
Exclusive white-glove Replica production
White-glove service to source, film, train, and license fully custom, exclusive AI replicas tailored to your needs
Early access to new models
Get first access to new models and products early and often
Pricing FAQs
Which plan is right for me?
Starter is built for developers to quickly start building with the Tavus product and APIs.
Growth is for developers and teams that are starting to scale, need additional features, and productionization matters.
Enterprise is for teams seeking partnership level support, volume discounts, full white labeling, and more. Talk to our sales team to learn more.
Growth is for developers and teams that are starting to scale, need additional features, and productionization matters.
Enterprise is for teams seeking partnership level support, volume discounts, full white labeling, and more. Talk to our sales team to learn more.
What is the difference between a Stock and Custom Replica?
Stock Replicas are a curated selection of replicas which have been created by the Tavus team, whereas Custom Replicas are digital twins of you or your users
How do concurrent conversation limits work?
The maximum amount of conversations you can have ongoing at any given time. For a higher limit than offered on self-serve plans, chat with sales.
How and when are minutes consumed?
Minutes are only consumed after the video or conversation has been generated.
Video Generation and CVI minutes are rounded up to the nearest ~6 second increment. For example, if your video output or conversation is 1:05, you will be billed for 1.1 minutes. Additionally, each incurs a minimum of a 30 second charge given GPU boot times.
Video Generation and CVI minutes are rounded up to the nearest ~6 second increment. For example, if your video output or conversation is 1:05, you will be billed for 1.1 minutes. Additionally, each incurs a minimum of a 30 second charge given GPU boot times.
How does pay-as-you-go work?
All Tavus plans include a certain amount of usage included, and additional usage is paid at the stated rate.
How to minimize CVI usage while testing?
When a conversation creation request is sent, a conversation is initiated and billed, regardless of if a participant joins. As a safeguard in case a conversation isn't properly closed, a conversation automatically closes after 5 minutes if a participant isn’t present, however these limits can be modified by setting the properties when creating a conversation
How do I upgrade?
Navigate to your billing page within the Tavus developer portal, and easily select the desired plan.
Do you offer custom pricing or plans?
Specific use cases may require custom plans, and our team is happy to discuss your specific goals. To talk to our sales team, you can sign up for a demo here.
Where can I see my usage and invoices?
Usage is displayed in the bottom left hand corner of your portal. Alternatively, navigate to your billing portal within the Tavus portal.
What does "white-labeled" mean?
Tavus provides fully white-labeled APIs which can be embedded into your product seamlessly, without Tavus branding. The exception is a required consent script a user must read which mentions Tavus when creating a custom replica.
Enterprise users are also able to white-label the consent passage.
Enterprise users are also able to white-label the consent passage.
What is included in Enterprise support?
Enterprise users have access to a private, dedicated Slack channel. We host a kick off meeting, and provide build support with API implementation, prompt engineering, integrations, customizations, UX/UI reviews, and more.
We're your internal AI video team- and that means we’re building together.
We're your internal AI video team- and that means we’re building together.
I have a question, how can I get in touch?
You can reach us at billing@tavus.io
The OS for human-AI interaction
Build immersive AI-generated video experiences in your application