confident-ai--deepeval/docs/content/integrations/index.mdx

---
id: integrations
title: Integrations Overview
sidebar_label: Overview
---

import { OpenAIMark } from "@site/src/components/BrandMarks";

DeepEval integrates with the frameworks, model providers, and data stores teams already use to build LLM applications. Use these pages to connect tracing, evaluation, synthetic data, and model configuration to your existing stack.

## Frameworks

Framework integrations let DeepEval evaluate entire execution traces without manually orchestrating every intermediate step. Use these when you want traces, spans, and component-level evals to line up with the framework your agents, chains, tools, and workflows already run on.

<Cards>
  <Card
    icon={
      <img
        src="/icons/integrations/langchain.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="LangChain"
    href="/integrations/frameworks/langchain"
    description="Trace and evaluate LangChain chains, tools, and agents."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/pydanticai.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="Pydantic AI"
    href="/integrations/frameworks/pydanticai"
    description="Trace Pydantic AI agents and evaluate their outputs."
  />
  <Card
    icon={<OpenAIMark width={20} height={20} />}
    title="OpenAI Agents"
    href="/integrations/frameworks/openai-agents"
    description="Evaluate workflows built with the OpenAI Agents SDK."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/langgraph.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="LangGraph"
    href="/integrations/frameworks/langgraph"
    description="Trace and evaluate graph-based agent workflows."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/agentcore.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="AgentCore"
    href="/integrations/frameworks/agentcore"
    description="Instrument AWS AgentCore agents with OpenTelemetry traces."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/strands.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="Strands"
    href="/integrations/frameworks/strands"
    description="Instrument Strands Agents SDK apps with OpenTelemetry traces."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/google-adk.png"
        alt=""
        width={20}
        height={20}
      />
    }
    title="Google ADK"
    href="/integrations/frameworks/google-adk"
    description="Trace Google ADK agents through OpenTelemetry and OpenInference."
  />
  <Card
    icon={
      <img
        src="/icons/integrations/llamaindex.svg"
        alt=""
        width={20}
        height={20}
      />
    }
    title="LlamaIndex"
    href="/integrations/frameworks/llamaindex"
    description="Instrument LlamaIndex retrieval and agent pipelines."
  />
  <Card
    icon={
      <img src="/icons/integrations/crewai.svg" alt="" width={20} height={20} />
    }
    title="CrewAI"
    href="/integrations/frameworks/crewai"
    description="Trace CrewAI crews, agents, tasks, and tool calls."
  />
  <Card
    icon={<OpenAIMark width={20} height={20} />}
    title="OpenAI"
    href="/integrations/frameworks/openai"
    description="Trace OpenAI SDK calls and evaluate OpenAI-powered apps."
  />
  <Card
    icon={
      <img src="/icons/integrations/claude.svg" alt="" width={20} height={20} />
    }
    title="Anthropic"
    href="/integrations/frameworks/anthropic"
    description="Trace Anthropic model calls inside DeepEval workflows."
  />
</Cards>

## Evaluation Models

Evaluation model integrations configure the LLM provider DeepEval uses for LLM-as-a-judge metrics, synthetic data generation, conversation simulation, and prompt optimization. Pick the provider that matches your infrastructure, latency, privacy, and cost needs.

<Cards className="md:grid-cols-3">
  <Card title="OpenAI" href="/integrations/models/openai" />
  <Card title="Azure OpenAI" href="/integrations/models/azure-openai" />
  <Card title="Ollama" href="/integrations/models/ollama" />
  <Card title="OpenRouter" href="/integrations/models/openrouter" />
  <Card title="Anthropic" href="/integrations/models/anthropic" />
  <Card title="Amazon Bedrock" href="/integrations/models/amazon-bedrock" />
  <Card title="Gemini" href="/integrations/models/gemini" />
  <Card title="DeepSeek" href="/integrations/models/deepseek" />
  <Card title="Vertex AI" href="/integrations/models/vertex-ai" />
  <Card title="Grok" href="/integrations/models/grok" />
  <Card title="Moonshot" href="/integrations/models/moonshot" />
  <Card title="Portkey" href="/integrations/models/portkey" />
  <Card title="vLLM" href="/integrations/models/vllm" />
  <Card title="LM Studio" href="/integrations/models/lmstudio" />
  <Card title="LiteLLM" href="/integrations/models/litellm" />
</Cards>

## Vector DBs

Vector database integrations show how to connect retrieval systems to DeepEval so RAG metrics can evaluate the context your application actually retrieves. Use these examples to benchmark retrieval quality and end-to-end RAG behavior.

<Cards>
  <Card title="Cognee" href="/integrations/vector-databases/cognee" />
  <Card
    title="Elasticsearch"
    href="/integrations/vector-databases/elasticsearch"
  />
  <Card title="Chroma" href="/integrations/vector-databases/chroma" />
  <Card title="Weaviate" href="/integrations/vector-databases/weaviate" />
  <Card title="Qdrant" href="/integrations/vector-databases/qdrant" />
  <Card title="PGVector" href="/integrations/vector-databases/pgvector" />
</Cards>

## Others

Integrations that don't fit cleanly into the categories above — typically training/eval-time hooks rather than runtime tracing.

<Cards>
  <Card
    title="Hugging Face"
    href="/integrations/frameworks/huggingface"
    description="Run DeepEval callbacks during Hugging Face training/evaluation."
  />
</Cards>