ModelRiver Documentation
A comprehensive guide to building AI-native infrastructure with ModelRiver. Explore all platform features, SDKs, and integration patterns.
Build a real-time AI chatbot
Experience end-to-end streaming AI that most AI gateways don't offer. Use our core SDK for instant WebSocket delivery and CLI for local development.
Get started
Step-by-step guidance to set up and connect your first project in minutes.
Quick answers to commonly asked questions about platform features.
Platform
Visual node-based orchestration for complex, multi-step AI pipelines.
Send requests to the unified ModelRiver endpoint, use OpenAI-compatible SDKs, handle streaming, and more.
Build event-driven AI architectures with async endpoints, webhook callbacks, and real-time delivery to clients.
Manage your projects, monitor usage, and configure settings.
Manage authentication credentials for programmatic access.
Define and enforce specific JSON schemas for AI responses.
Reduce latency and costs by caching frequent AI responses.
Real-time event subscriptions to trigger external automation.
Monitor, debug, and analyze every AI request with comprehensive logging, timelines, and webhook tracking.
Framework-specific libraries for React, Python, Node.js, and more.
Enterprise-grade RBAC, data encryption, and transparent audit logging.
Diagnostics and solutions for common challenges.
Connect ModelRiver to LLM frameworks, agent systems, backend stacks, knowledge and memory, and automation tools.
Solutions
Build production-grade AI applications with automatic model failover and redundancy.
Everything you need to know about implementing real-time AI streaming using WebSockets.
Stop parsing text and start using JSON with validated Structured Outputs.