MetaHuman AI for live conversations.
Connect your MetaHuman to IAMX so people can speak or type to the character and receive spoken replies. Keep the character, scene and animation setup in Unreal Engine; manage its instructions, voice, knowledge and configured tools in the IAMX panel.
Connect an Unreal character to conversational AI
- Create the IAMX character
Choose the voice and AI provider, write the character instructions and configure the tools or knowledge it can use.
- Connect the Unreal plugin
Install a compatible plugin build, add the IAMX component and configure the character connection. Follow the installation and animation-node guides for your project.
- Route speech and animation
Use the plugin’s speech and facial animation flow for your talking NPC. Trigger your own game or training-scene behavior through configured actions.
- Choose local rendering or streaming
Render on a kiosk or game machine, or use Pixel Streaming to deliver the Unreal scene to a browser.
GLB and MetaHuman are different rendering paths
| Requirement | Browser GLB | Unreal MetaHuman |
|---|---|---|
| Rendering | On the visitor’s device in the browser. | In an Unreal application on a local or remote rendering machine. |
| Website delivery | Native 3D character in the widget. | Pixel Streaming for the rendered Unreal scene. |
| Transparent widget | Supported for the character-only GLB template. | Use the streamed scene or a portrait frame. |
| Animation setup | Compatible GLB skeleton and morph targets. | Unreal character rig, animation setup and IAMX integration. |
Plan the deployment around your scene
Rendering cost and responsiveness depend on the scene, hardware, network, AI provider and concurrent sessions. A streamed MetaHuman requires suitable rendering capacity as well as the conversation service. Test the complete deployment before choosing capacity or promising latency.
Keep scene control in Unreal, configure conversation in IAMX
In the IAMX panel
Create the character, choose the conversation model and voice, write its role and boundaries, and enable the knowledge files or tools it should use. Test representative questions before connecting the final scene.
In your Unreal project
Install the plugin build that matches your engine version, add the IAMX component and connect the character. Configure the animation nodes for the intended face and gaze behavior, then wire the Blueprint events your scene needs.
For an AI NPC, a training simulation or a virtual-world guide, begin with one complete interaction: a visitor asks a question, the character answers and a configured action affects the scene. Sales roleplay or a virtual showroom uses the same connection; your own scene logic determines each action.
What to measure in a MetaHuman pilot
| Check | Use your intended deployment |
|---|---|
| Speech and face | Try the actual voice, languages and target character. Watch mouth timing, listening transitions and facial motion across short and long replies. |
| Conversation quality | Use approved source questions, ambiguous requests and requests outside the character's scope. Verify the enabled knowledge and tool behavior. |
| Time to a reply | Measure the complete path from microphone input to audible response. Model, speech service, network and rendering each contribute. |
| Browser delivery | For Pixel Streaming, test the signaling connection, microphone permissions, video playback and recovery after an interrupted network connection. |
| Capacity and cost | Evaluate your scene on the selected rendering hardware at the intended concurrency. Budget for conversation usage, any separately billed providers and rendering infrastructure. |
IAMX conversation plans do not include cloud-rendering minutes. A character allowance is not a guarantee of the same number of simultaneous streamed sessions. See plans and quota before sizing the deployment.
MetaHuman AI FAQs
Does a visitor need to install Unreal Engine?
For a Pixel Streaming deployment, the visitor uses a compatible browser while Unreal runs on your rendering host. A locally rendered Unreal application runs on the machine where you deploy that application.
Can a MetaHuman float transparently over my website?
The streamed video includes the rendered background. Use a portrait or scene frame for MetaHumans. The transparent character-only widget uses the browser GLB path.
Can it answer using company documents?
Enable the relevant knowledge files for the IAMX character on a plan with knowledge access. Validate retrieved passages and spoken answers with your own material before using the character publicly.
Does this convert a MetaHuman into a browser GLB?
No. This integration connects an Unreal character to IAMX conversation. Browser GLB avatars have their own model, rig and animation requirements; switching presentation does not convert the Unreal rig.
Build a character people can talk to.
Start with a model, give it a voice and knowledge, and test the complete experience before publishing.
Maintained by IAMX · Updated · Platform facts and requirements