AI with character.
- BoltExpression 2
- MariEssence 2
- Chef WaddlesExpression 2
- Shelly TidewaterExpression 2
- KwameEssence 2
- Captain AxlExpression 2
Photoreal people. Curious creatures. A character that feels like yours.
What would you like to build?
Start with the experience you want. Choose the model and rendering location along the way.
01For your existing stack
Give your agent a face
Add a real-time avatar to your voice agent or app. Keep the conversation services you already use.
02For your space or device
Put a character on site
Render on your own hardware for a kiosk, an app, or an embedded screen. Standard sessions connect online; offline requires a separate license.
03For a managed experience
Create a hosted agent
Build a character with a voice and personality, then share its link or embed it. bitHuman hosts the managed conversation.
Two models. Endless personalities.
The avatar model turns speech into a face. Your conversation stack gives it something to say.
Photoreal people
Essence 2
Turn one portrait into a lifelike person, lip-synced to speech in real time.
Meet Essence 2
Any character
Expression 2
Bring a person, animal, mascot, or object to life from one image.
Meet Expression 2Essence 1 and Expression 1 remain maintained first-generation models. Expression 1 is cloud-only. First-generation guide
Choose where it comes to life.
Rendering location, conversation services, and network access are separate choices.
bitHuman cloud
Online rendering on bitHuman’s servers. Bring your own conversation services for avatar-only use, or choose bitHuman’s managed conversation.
Your servers
Rendering on your machines, with voice and language services you choose. Standard sessions authenticate and report usage online.
On the device
Rendering on supported hardware or WebGPU browsers. Conversation services are separate; standard sessions authenticate and report usage online.
Fully offline
Licensed Essence 2 and Expression 2 rendering on Linux x86_64, off the internet once set up. The conversation pipeline is agreed with sales.
Hardware, browser, and model requirements apply. Check supported platforms before choosing on-device rendering.
Local rendering, connected services
When the avatar renders on your hardware and the conversation runs on your own services, bitHuman receives usage metering only — never audio, video or conversation text.
With the web embed, the conversation runs on bitHuman's servers, even when the avatar renders in the tab. The local brain still signs in and reports usage online, so it needs a network connection; it is not part of the offline license.
Trace the data flowA dedicated offline license
Essence 2 and Expression 2 run fully offline on Linux x86_64 under an offline license for Business and Enterprise. Creation still happens online. Under an offline license, where the conversation runs is agreed with sales for your site.
Check offline requirementsCharacters out in the world.
Real deployments at museums, trade shows, and live events.
DeploymentMINT Museum of Toys
Multilingual guidance and ticketing from The Little Prince.
DeploymentIvoclar Vivadent
Virtual Einstein answers visitors at trade shows.
DeploymentNRF: Retail’s Big Show
Avatar kiosk guides help attendees find their way.
Demos built by bitHuman
See what a shopping or checkout experience can look like.
Creation is separate from running time: it happens online and takes hours. Expression 2 creation consumes the Creator plan’s monthly credit grant.
For avatar-only use, you supply the conversation services. Managed conversation adds bitHuman-hosted voice and language services. Active session time counts while talking or idle; check the current pricing and billing guide for rates.
Looking for the consumer app? bitHuman Live is coming soon.