Integrations

AdalFlow integrates with many popular AI and database platforms to provide a comprehensive solution for your LLM applications.

Model Providers

AdalFlow supports a wide range of model providers, each offering unique capabilities and models:

Deepseek

Anthropic

Mistral

Ollama

OpenAI

Transformers

Groq

Azure OpenAI

Amazon Bedrock

LLM + VLLM

Major Class	Description	Tutorial
Generator	A user-facing orchestration component that handles LLM predictions. It includes a prompt template, model client, and output parser.	Generator
ReActAgent	An agent that uses large language model reasoning (Re) and actions (Act) to solve queries.	Agent
ModelClient	The low-level component managing the actual calls to a chosen LLM (OpenAI, Anthropic, VLLM, etc.).	ModelClient

Vector Databases

Qdrant

LanceDB

Embedding and Reranking Models

Hugging Face

OpenAI

Cohere

Embeddings, Reranking, and Vector Databases

Major Class	Description	Tutorial
Embedder	A user-facing component that orchestrates embedding models via ModelClient and output_processors	Embedder
Retriever	Each subclass can be a local, a vector-db, a retranker, or an LLM-turned retriever to handle retrieval tasks in RAG.	Retriever
TextSplitter	Chunking large text into smaller segments for more efficient and accurate embedding, retrieval, and LLM context processing.	TextSplitter