Zep logo

Zep

Zep is a context engineering and agent memory platform that assembles relevant context from chat history, business data, and user interactions for AI agents. It builds a temporal knowledge graph per user that evolves as new facts arrive, automatically extracts entities, and invalidates facts that are superseded. Zep exposes a Graph RAG layer with customizable context blocks for fast personalization. The platform offers Cloud and self-hosted community editions with Python, TypeScript, and Go SDKs and integrations with LangChain, LangGraph, LlamaIndex, CrewAI, and AutoGen. Zep targets sub-200ms p95 retrieval latency and is SOC 2 Type II and HIPAA aligned for enterprise deployments.

1 APIs 0 Features
MemoryAI AgentsKnowledge GraphTemporal GraphGraph RAGContext EngineeringLangChainLlamaIndexCrewAILLMsPersonalizationRetrieval

Zep publishes 1 API on the APIs.io network. Tagged areas include Memory, AI Agents, Knowledge Graph, Temporal Graph, and Graph RAG.

Zep’s developer surface includes documentation, engineering blog, pricing, and 7 more developer resources.

APIs

Zep Cloud API

The Zep Cloud API delivers agent memory and temporal knowledge graph services over REST. It exposes endpoints for users, sessions, messages, memory retrieval, the per-user graph...

Resources

🔗
Website
Website
🔗
Documentation
Documentation
📰
Blog
Blog
👥
GitHubOrganization
GitHubOrganization
💰
Pricing
Pricing
📜
TermsOfService
TermsOfService
📜
PrivacyPolicy
PrivacyPolicy
🔗
Discord
Discord
🔗
X
X
🔗
LinkedIn
LinkedIn

Sources

apis.yml Raw ↑
aid: zep
name: Zep
description: >-
  Zep is a context engineering and agent memory platform that assembles
  relevant context from chat history, business data, and user interactions for
  AI agents. It builds a temporal knowledge graph per user that evolves as new
  facts arrive, automatically extracts entities, and invalidates facts that
  are superseded. Zep exposes a Graph RAG layer with customizable context
  blocks for fast personalization. The platform offers Cloud and self-hosted
  community editions with Python, TypeScript, and Go SDKs and integrations
  with LangChain, LangGraph, LlamaIndex, CrewAI, and AutoGen. Zep targets
  sub-200ms p95 retrieval latency and is SOC 2 Type II and HIPAA aligned for
  enterprise deployments.
type: Index
position: Provider
access: 3rd-Party
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
tags:
  - Memory
  - AI Agents
  - Knowledge Graph
  - Temporal Graph
  - Graph RAG
  - Context Engineering
  - LangChain
  - LlamaIndex
  - CrewAI
  - LLMs
  - Personalization
  - Retrieval
url: https://raw.githubusercontent.com/api-evangelist/zep/refs/heads/main/apis.yml
created: '2026-05-23'
modified: '2026-05-23'
specificationVersion: '0.20'
apis:
  - aid: zep:cloud-api
    name: Zep Cloud API
    description: >-
      The Zep Cloud API delivers agent memory and temporal knowledge graph
      services over REST. It exposes endpoints for users, sessions, messages,
      memory retrieval, the per-user graph, facts, summaries, and customizable
      context blocks for prompt assembly.
    humanURL: https://help.getzep.com
    baseURL: https://api.getzep.com/api/v2
    tags:
      - Memory
      - Graph
      - Sessions
      - Users
      - Facts
      - Context
      - REST
    properties:
      - type: Documentation
        url: https://help.getzep.com
      - type: GettingStarted
        url: https://help.getzep.com/v2/quickstart
      - type: SignUp
        url: https://app.getzep.com
      - type: SDK
        url: https://github.com/getzep/zep-python
      - type: SDK
        url: https://github.com/getzep/zep-js
      - type: SDK
        url: https://github.com/getzep/zep-go
      - type: GitHubRepository
        url: https://github.com/getzep/zep
      - type: Pricing
        url: https://www.getzep.com/pricing
      - type: Authentication
        url: https://help.getzep.com/projects
    features:
      - name: Temporal Knowledge Graph
        description: Per-user knowledge graph that captures entities, relationships, and time-aware facts.
      - name: Memory and Sessions
        description: Manage users and chat sessions, with messages persisted for retrieval and summarization.
      - name: Automatic Fact Extraction
        description: Extract structured facts from messages and invalidate facts that are superseded.
      - name: Graph RAG
        description: Retrieve graph-grounded context for prompts using semantic and graph traversal queries.
      - name: Customizable Context Blocks
        description: Compose reusable, customizable blocks of context to inject into agent prompts.
      - name: Custom Entity Types
        description: Define domain-specific entity types to capture vertical-specific knowledge.
      - name: Multi-Source Ingestion
        description: Ingest from chat, documents, and structured JSON to enrich the graph.
      - name: Low-Latency Retrieval
        description: Sub-200ms p95 retrieval latency for use in real-time agent loops.
      - name: Enterprise Compliance
        description: SOC 2 Type II and HIPAA-aligned controls for regulated workloads.
    useCases:
      - name: Personal AI Assistants
        description: Maintain long-term user memory across conversations and devices.
      - name: Customer Support Copilots
        description: Provide agents with persistent customer history and account facts.
      - name: Sales Agents
        description: Track deal context, stakeholders, and prior interactions per account.
      - name: Healthcare Workflows
        description: Persist patient context with HIPAA-aligned controls and audit trails.
      - name: Voice Agents
        description: Inject low-latency context blocks into real-time voice agent prompts.
    integrations:
      - name: LangChain
      - name: LangGraph
      - name: LlamaIndex
      - name: CrewAI
      - name: AutoGen
      - name: OpenAI
      - name: Anthropic
      - name: Vercel AI SDK
      - name: Pipecat
      - name: LiveKit
    authentication:
      - type: API Key
        description: Zep Cloud uses an Api-Key header with project keys prefixed with z_; self-hosted uses Bearer tokens.
common:
  - type: Website
    url: https://www.getzep.com
  - type: Documentation
    url: https://help.getzep.com
  - type: Blog
    url: https://blog.getzep.com
  - type: GitHubOrganization
    url: https://github.com/getzep
  - type: Pricing
    url: https://www.getzep.com/pricing
  - type: TermsOfService
    url: https://www.getzep.com/terms-of-service
  - type: PrivacyPolicy
    url: https://www.getzep.com/privacy-policy
  - type: Discord
    url: https://discord.com/invite/W8Kw6bsgXQ
  - type: X
    url: https://x.com/zep_ai
  - type: LinkedIn
    url: https://www.linkedin.com/company/zep-ai
maintainers:
  - FN: Kin Lane
    email: [email protected]