Integrate ModelRiver everywhere

Why integrate with ModelRiver?

ModelRiver works behind any tool that can call an OpenAI-compatible endpoint. Point your framework or SDK at https://api.modelriver.com/v1, set your ModelRiver API key, and you instantly get:

• Automatic failover across multiple AI providers
• Cost tracking and token-level analytics
• Structured outputs enforced at the workflow level
• Real-time observability for every request
• Provider switching without code changes

LLM frameworks

Orchestration libraries for building LLM-powered pipelines, RAG systems, and multi-step AI workflows.

LangChain

Python's most popular LLM orchestration framework. Use ModelRiver as a drop-in ChatOpenAI replacement for chains, agents, RAG pipelines, and tool-calling workflows.

LlamaIndex

Data-aware LLM framework for RAG, knowledge graphs, and document QA. Swap in ModelRiver's endpoint to get automatic failover across every index query.

Haystack

Deepset's production-ready NLP framework. Connect Haystack generators and pipelines to ModelRiver for resilient, observable AI operations.

Agent frameworks

Multi-agent and autonomous AI systems powered by ModelRiver's resilient routing.

AutoGen

Microsoft's multi-agent conversation framework. Wire ModelRiver into AutoGen's chat-based agent topology for resilient, multi-provider autonomous conversations.

CrewAI

Multi-agent orchestration framework where each crew member is an AI agent. Give every agent a ModelRiver workflow with its own model, fallbacks, and structured output schema.

OpenAI Agents SDK

OpenAI's official agent framework with built-in tool use, handoffs, and guardrails. Connect it to ModelRiver for provider-agnostic agent execution with automatic failover.

LangGraph

LangChain's graph-based agent framework for building stateful, multi-step agent workflows with cycles, branching, and persistence.

Backend frameworks

End-to-end guides for embedding ModelRiver in your web application backend.

Phoenix

Elixir's real-time web framework. Build AI-powered LiveView applications with ModelRiver channels, webhooks, and streaming responses.

Next.js

Full-stack React framework by Vercel. Use the Vercel AI SDK with ModelRiver for streaming chat UIs, server actions, and edge functions.

FastAPI

High-performance Python API framework. Build AI-powered REST and WebSocket endpoints backed by ModelRiver.

Django

Python's batteries-included web framework. Add AI features to Django views, REST APIs, and Celery tasks with ModelRiver.

Knowledge and memory

Connect ModelRiver embeddings to your vector store for RAG, semantic search, and similarity matching.

Pinecone

Fully managed vector database. Generate embeddings through ModelRiver workflows and store them in Pinecone for production-scale semantic search.

Weaviate

Open-source vector search engine. Use ModelRiver as the vectoriser for Weaviate collections with automatic provider failover.

Supabase

Open-source Firebase alternative with pgvector support. Store ModelRiver-generated embeddings in Supabase for full-stack AI applications.

Observability

Monitor and debug ModelRiver requests alongside your application telemetry.

Datadog

Enterprise observability platform. Forward ModelRiver request metrics, traces, and logs to Datadog for unified monitoring dashboards.

Sentry

Error tracking and performance monitoring. Capture ModelRiver errors, slow requests, and provider failover events in Sentry.

Automation

Connect ModelRiver to thousands of apps without writing code.

Zapier

Automate AI-powered workflows across 6,000+ apps. Trigger ModelRiver requests from form submissions, Slack messages, CRM updates, and more.

Quick reference

Next steps

• API reference: Full endpoint documentation
• Workflows: Configure providers, fallbacks, and structured outputs
• Observability: Monitor every integration request