Direct Mode
AvatarKit connects from the client to Motion Server using a short-lived Session Token. The client renders locally, sends audio, receives motion data, and owns its lifecycle.
Architecture comparison
Spatius Direct Mode vs Backend Mode
Choose Direct Mode when the client can connect directly to Motion Server with a Session Token and send existing speech audio. Choose Backend Mode when your server must own the Motion Server connection, payload routing, transport, and centralized session control.
Verified Aug 3, 2026Decision-ready guidePrimary sources
Decision matrix
Compare the complete path.
Do not compare isolated numbers unless the definitions, inputs, environment, and included services match.
| Decision area | Direct Mode | Backend Mode |
|---|---|---|
| Motion Server connection | AvatarKit client | Customer backend Server SDK |
| Credential at runtime | Session Token on client | Backend credentials and server session |
| Downstream transport | Direct Motion Server path | LiveKit or customer-owned transport |
| Client role | Send audio, receive motion, render | Receive backend payloads and render |
| Server control | Token endpoint and application services | Session, routing, payload, and egress control |
| Implementation effort | Shorter path | More infrastructure and flexibility |
| Best fit | Existing client-side speech audio | Backend-owned production pipeline |
How it works
Two different operating models.
Architecture decides which team owns rendering, transport, recovery, and the surrounding AI product.
Backend Mode
A Server SDK connects from the customer backend to Motion Server. The backend forwards encoded audio and motion payloads to clients using LiveKit or a customer-owned transport.
Best fit
Choose for the system you can operate.
The best option is the one whose responsibilities match your product, client, network, and team.
Choose Direct Mode when…
- Need the shortest standalone path
- Client already has speech audio
- Session Token endpoint is available
- No custom backend media transport is required
Choose Backend Mode when…
- Backend owns ASR/LLM/TTS pipeline
- Need central routing and observability
- Use LiveKit as downstream transport
- Custom payload transport is required
Limitations and unknowns
Direct Mode still requires a secure token endpoint; API keys must not be shipped to clients. Backend Mode adds operational responsibility and should not be chosen only because it sounds more enterprise-ready.
Build a controlled evaluation.
Use one workload and record both user experience and operational responsibility.
1. Evaluation stepIdentify who creates and times speech audio.
2. Evaluation stepDecide where the Motion Server session should live.
3. Evaluation stepChoose downstream transport and recovery ownership.
4. Evaluation stepMap credentials before building the client.
- Define the user job and acceptable fallback.
- Use the same input, session duration, and client.
- Record latency, traffic, compute, errors, and recovery.
- Compare total operating cost, not only list price.
Evidence
Sources and freshness.
Last verified Aug 3, 2026. Recheck implementation details when SDKs or plan terms change.
Related decisions