Architecture overview

Choose the right AI request pattern

ModelRiver provides three distinct ways to handle your AI traffic, designed to match the specific needs of your application's architecture. Compare their capabilities below, or dive into their interactive flow diagrams.

Detailed Comparison

A side-by-side break down of scaling properties and best use cases for each method.

CapabilityEvent-driven AsyncAsync WebhooksSynchronous
Difficulty LevelMediumModerateEasy
Scaling LevelAuto-scales infinitelyScales horizontallyLimited by connection threads
Best ForReal-time apps plus background/bulk tasks with proper webhook event notification.Background processing, bulk operations.Internal tools, quick one-off calls.
Latency Added~120ms (Socket handshake)< 50ms (Direct push)Absolute zero
Reliability against dropsPerfectPerfectVulnerable to timeouts
Delivery MethodWebSocket frames + REST Webhook execution dataDelivery to webhook and websocket simultaneouslyBlocking HTTP Response
Provider Retries / FailoversFully unblocked with automatic callback to ModelRiver and user notification.Robust background error handling.Handled in the single HTTP transaction window.