Fast Acknowledgment
Immediate response with a secure channel ID. No open connection holding, fundamentally eliminating gateway timeouts on slow generations.
Event-driven Async
Event-driven Architecture
Real-time updates via WebSockets with disconnected backend processes. Separation of concerns provides ultimate scalability.
1
Recommended
Event-driven async
Real-time updates via WebSocket. Backend callbacks for processing. Best for interactive apps.
Learn more2
Async (Dual-Delivery)
Fire and forget. Results delivered simultaneously to your webhook and browser. Best for reliable background ops.
Learn more3
Sync
Simple request/response. Blocks until complete. Best for quick operations.
Learn moreEvent-driven async flow
Step-by-step interactive data flow visualization
1
User prompt
Frontend interaction
2
Your backend
API call to MR
3
ModelRiver
Orchestration
4
AI providers
Auto-failover flow
Async workflow triggered
6
Your webhook
Callback to your server
Callback with final data to ModelRiver
5
ModelRiver
Message orchestration
7
User's browser
Real-time WS update
Architecture
Why event-driven async?
Separate ingestion from inference. ModelRiver's resilient infrastructure handles the operational complexity of AI requests, ensuring your application seamlessly scales.
Background Execution
AI runs asynchronously in isolated worker pools. Your backend remains lightning fast and completely insulated from thread exhaustion.
Automatic Fallback
Seamless multi-provider failover baked deeply into the pipeline. Achieve five-nines orchestration without altering any of your client architecture.
Instant Delivery
Real-time WebSocket conduits autonomously push the final streaming payloads directly back to the active client layer.