robots.txt
See which AI crawlers other sites block →# Content Signal preferences — https://contentsignals.org/
# search=yes allow AI/search indexing
# ai-input=yes allow use as input for AI answers (RAG/grounding)
# ai-train=yes allow use for AI model training
User-agent: GPTBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: OAI-SearchBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: ChatGPT-User
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: ClaudeBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Claude-SearchBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Claude-User
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: PerplexityBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Perplexity-User
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Google-Extended
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Googlebot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: CCBot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Bytespider
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Amazonbot
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: Applebot-Extended
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
User-agent: *
Content-Signal: search=yes, ai-input=yes, ai-train=yes
Allow: /
Disallow: /private/
Disallow: /admin/
Disallow: /api/
Disallow: /dev/
Sitemap: https://www.getzep.com/sitemap.xml
Sitemap: https://www.getzep.com/a/sitemap.xml
llms.txt
Compare with other sites' llms.txt files →# Zep
> Zep is the unified context layer for enterprise data. It connects business data, documents, and conversations into governed context graphs that give agents the context they need to complete tasks correctly.
Zep provides unified, governed context for production AI agents: a Context Lake of context graphs, sub-200ms retrieval at scale, and SOC 2 Type II + HIPAA compliance. Deploy in Zep's cloud, your VPC, or with your own keys.
Built on the open-source Graphiti framework and powered by Konig, Zep's proprietary graph database service. SDKs for Python, TypeScript, and Go.
Benchmarked on LoCoMo and LongMemEval against mem0 and other agent-memory systems.
## Platform
- [Context Lake | Governed context for agents | Zep](https://www.getzep.com/platform/context-lake/): The Context Lake is the unified context layer for enterprise data, giving every agent governed, up-to-date facts linked to their sources.
- [Konig — Zep's proprietary graph database service](https://www.getzep.com/platform/agent-knowledge-graph/): Konig is Zep's graph database service and the data plane beneath its unified context layer. Millions of governed graphs, served in sub-200ms.
- [Graphiti](https://www.getzep.com/platform/graphiti/): Graphiti is the open-source context graph framework that powers Zep.
## Product
- [Agent memory | Solutions built with Zep](https://www.getzep.com/product/agent-memory/): Combine conversations, business data, and user activity into context that agents can retrieve for later tasks.
## Solutions
- [Company Brain for Internal Agents](https://www.getzep.com/solutions/company-brain/): Give agents one governed company brain with Zep, connecting projects, policies, and decisions using source-linked facts.
- [Customer Context for Product Agents](https://www.getzep.com/solutions/customer-context/): Give the agents in your product current, unified context on each customer with Zep, built from conversations and account records.
- [Vertical Agents for Regulated Industries](https://www.getzep.com/solutions/vertical-agents/): Build legal, finance, and healthcare agents on Zep. Each customer gets its own graph, and every request is logged.
- [Context Engineering for AI Agents](https://www.getzep.com/solutions/context-engineering/): Context engineering is the practice of designing and managing the context AI agents see at every call. Unified context is what agents need. The Context Lake is how Zep delivers it.
- [The Mem0 alternative — agent memory on Zep, built for production](https://www.getzep.com/mem0-alternative/): Zep is the unified context layer for production agents. Higher accuracy, faster retrieval, fewer tokens than mem0 — plus the customization, governance, and scale that agents need.
## Research
- [Research](https://www.getzep.com/research/): Research papers from the Zep team on temporal context graphs and on how agents learn skills from their own runs.
- [Agent memory benchmarks](https://www.getzep.com/research/memory-benchmarks/): Benchmark results on LoCoMo and LongMemEval, two industry benchmarks for long-running agent memory.
- [S&P Global Market Intelligence: Zep Tackles Agent Memory Limitations Through Its Temporal Context Graph](https://www.getzep.com/research/sp-global-market-intelligence-zep-coverage/): S&P Global Market Intelligence initiates analyst coverage on Zep. Read excerpts and download the full report.
- [S&P Global Market Intelligence on Zep](https://www.getzep.com/analysts/sp-market-intelligence-report/): Read the S&P Global Market Intelligence report on Zep — analyst coverage of Zep's unified context layer and temporal context graph.
## Blog
- [Zep Blog: Context Engineering and Agent Memory](https://www.getzep.com/blog/): Engineering and product writing from Zep on context engineering, agent memory, context graphs and retrieval for production AI agents.
- [How Zep Context Graphs keep two timelines for every fact](https://www.getzep.com/blog/context-graphs-two-timelines-every-fact/index.md): When a fact changes, a Zep Context Graph keeps the old one with an end date. Agents can see what was true on any date, and when each fact was added.
- [Metadata is the easy part of a data context layer](https://www.getzep.com/blog/metadata-easy-part-data-context-layer/index.md): Data agents rarely fail at SQL. They fail at choosing the right table, metric or join key. We read 15 engineering write-ups to see how teams fix it.
- [7 reasons context graphs are unusually effective](https://www.getzep.com/blog/context-graphs-unusually-effective/index.md): Context graphs store entities and the relationships between them. Seven reasons explain why this structure gives AI agents better context than text chunks.
- [Everything we shipped in the Zep Context API](https://www.getzep.com/blog/what-shipped-in-the-context-api/index.md): Entity type hierarchies, the graph assistant, document context, identity properties, event time, and more. What each new Zep feature does and how to use it.
- [Hyperedges: accurate answers when a fact has many entities](https://www.getzep.com/blog/hyperedges-facts-across-many-entities/index.md): Zep v4 adds hyperedges. When one fact relates three or more entities, a search that matches one of them returns the full fact, with every related entity.
- [Zep v4: the Context API for agents](https://www.getzep.com/blog/zep-api-for-agents/index.md): Zep v4 is the Context API for agents: graph navigation, Ingestion Traces and stable error codes in the API, plus a step-by-step path from v3.
- [Defending Agent Memory Against Poisoning](https://www.getzep.com/blog/defending-agent-memory-poisoning/index.md): How memory poisoning attacks agent memory, the six controls that contain it, and what belongs in your app versus your memory layer.
- [Why we built a graph database service for agent memory](https://www.getzep.com/blog/why-we-built-a-graph-database-service-for-agent-memory/index.md): Konig serves millions of governed knowledge graphs, with p95 retrieval under 100ms.
- [Memory for every agent your team uses](https://www.getzep.com/blog/agent-memory-mcp/index.md): The Zep Memory MCP Server: agent memory for Claude, ChatGPT, and your custom agents, secured by your identity provider.
- [Coding agents can design your Zep implementation](https://www.getzep.com/blog/coding-agents-can-design-your-zep-implementation/index.md): Zep's plugin for Claude Code, Codex, and Cursor gives your coding agent a skill for designing agent memory implementations, plus live docs search.
- [Evaluating Nemotron 3 Embed for agent memory](https://www.getzep.com/blog/evaluating-nemotron-3-embed-for-agent-memory/index.md): NVIDIA Nemotron 3 Embed 1B took first place on every task in Zep's agent memory recall benchmark of 5,954 production queries.
- [How Zep tracks provenance in agent memory](https://www.getzep.com/blog/how-zep-tracks-provenance-in-agent-memory/index.md): An LLM synthesizes facts from many sources, erasing the source trail. How Zep tracks provenance in agent memory so every fact is traceable.
- [Securing Agent Memory with ABAC](https://www.getzep.com/blog/attribute-based-access-control/index.md): Attribute-Based Access Control scopes any Zep API key to the exact actions and agent memory it can access. Least privilege for every agent.
- [Unified agent memory in any MCP client](https://www.getzep.com/blog/unified-agent-memory-in-any-mcp-client/index.md): Unified agent memory in any MCP client. Your users' desktop, coding, and custom agents share the same governed memory as the agents you build.
- [Agent memory placement can cut your token bill up to 2x](https://www.getzep.com/blog/where-you-put-memory-in-the-prompt-can-cut-your-token-bill-up-to-2x/index.md): Agent memory must refresh every turn, but in the system prompt it breaks prompt caching. Moving it past the cache breakpoint cuts token costs up to 2x or more.
- [Markdown is not agent memory](https://www.getzep.com/blog/markdown-is-not-agent-memory/index.md): Markdown files work as agent memory for a single agent and user. At scale, as facts change, and across agents, the pattern breaks.
- [Building Agents in Go Without a Framework](https://www.getzep.com/blog/agentic-development-in-go/index.md): Why Go fits the agent runtime, whether you need a framework, and how to build the agent harness in Go on native vendor SDKs with agent memory.
- [Sycophancy is a design choice](https://www.getzep.com/blog/sycophancy-is-a-design-choice/index.md): Writer's paper says agent memory amplifies sycophancy up to 25x. Its own data traces the effect to experiment design. A teardown.
- [The Batch API: Load Large Datasets into Agent Memory](https://www.getzep.com/blog/the-batch-api-load-large-datasets/index.md): Zep's Batch API loads large datasets into agent memory faster, in batches up to 50,000 items, with a progress dashboard and no impact on real-time ingestion.
- [Smart Context Assembly: Fewer Tokens, Better Quality](https://www.getzep.com/blog/smart-context-assembly-fewer-tokens-higher-quality/index.md): Today we're announcing Smart Context Assembly, an upgrade to how Zep's default Context Block is built: higher accuracy from fewer tokens, with no code changes.
- [Observations: Patterns and Insights from the Context Graph](https://www.getzep.com/blog/observations-durable-patterns-from-the-context-graph/index.md): Observations are a new context type in Zep that capture patterns and insights across your Context Graphs, automatically discovered and surfaced to agents.
- [Zep's 5 Context Types: How to Use and Combine Each One](https://www.getzep.com/blog/zep-context-types/index.md): Zep produces five distinct types of context from a user's graph. Each captures something different.
- [Context You Can Trace, Filter, and Trust](https://www.getzep.com/blog/context-you-can-trace-filter-and-trust/index.md): Every fact in your agent's context graph came from somewhere.
- [Stop Letting Your Agent Decide What It Needs to Know](https://www.getzep.com/blog/stop-letting-your-agent-decide-what-it-needs-to-know/index.md): Your agent has the tools. It just doesn't call them — and smarter models won't fix that.
- [3 Decisions That Shape Every Agent's Context Architecture](https://www.getzep.com/blog/three-decisions-that-shape-every-agents-context-architecture/index.md): Every agent context architecture comes down to three decisions: scope, data sources, and retrieval strategy.
- [Evaluation and Control: Evaluation Framework, zepctl CLI, and Dashboard Overhaul](https://www.getzep.com/blog/evaluation-and-control-evaluation-framework-zepctl-cli-and-dashboard-overhaul/index.md): An evaluation framework for testing Zep against your data, a redesigned dashboard with analytics, and zepctl — a CLI for administering Zep projects.
- [Build Better Context Graphs: Custom Instructions, Search Filters, and Webhooks](https://www.getzep.com/blog/build-better-context-graphs-custom-instructions-search-filters-and-webhooks/index.md): Custom extraction instructions, property-level search filters, exclusion filters, and webhooks — more control over how you build and query context graphs.
- [Context Templates: Context Engineering Made Simple](https://www.getzep.com/blog/context-templates-context-engineering-made-simple/index.md): You're tuning an agent. Retrieve too little context and it hallucinates.
- [The Retrieval Tradeoff: What 50 Experiments Taught Us About Context Engineering](https://www.getzep.com/blog/the-retrieval-tradeoff-what-50-experiments-taught-us-about-context-engineering/index.md): Zep builds temporal knowledge graphs from conversations and business data, then automatically retrieves relevant context when your agent needs it.
- [How Zep Works: A Visual Guide to Knowledge Graphs for AI Agents](https://www.getzep.com/blog/a-visual-guide-to-knowledge-graphs-for-ai-agents/index.md): Agents don't fail because the model is bad. They fail because they don't have the right context.
- [Building Voice Agents with Memory: Zep x LiveKit](https://www.getzep.com/blog/zep-livekit/index.md): Create personalized voice agents with long-term memory with minimal added latency
- [Agents That Always Remember What Matters](https://www.getzep.com/blog/agents-that-always-remember-what-matters/index.md): Zep now offers steerable user summaries. This post explores why these are useful and how best to implement them.
- [Graphiti Hits 20K Stars! + MCP Server 1.0](https://www.getzep.com/blog/graphiti-hits-20k-stars-mcp-server-1-0/index.md): Graphiti crossed 20,000 GitHub stars today! Thanks for building with us.
- [How We Scaled Zep 30x in 2 Weeks (and Made It Faster)](https://www.getzep.com/blog/scaling-agent-memory-zep-30x/index.md): Our infrastructure broke under 30x growth. Six weeks later, we made Zep faster than before: 10x better latency, 92% faster processing.
- [Zep v3: Context Engineering Takes Center Stage](https://www.getzep.com/blog/zep-v3-context-engineering-takes-center-stage/index.md): Context engineering > prompt engineering: Zep v3 assembles memory & business data for agents that work.
- [Graphiti Adds FalkorDB Support as Project Approaches 14,000 Stars](https://www.getzep.com/blog/graphiti-knowledge-graphs-falkordb-support/index.md): FalkorDB contributes integration to Graphiti. Project approaching 14K stars & 25K weekly PyPI downloads.
- [What is Context Engineering, Anyway?](https://www.getzep.com/blog/what-is-context-engineering/index.md): From Prompt Engineering to Context Engineering: The Why's and How.
- [The Private Agent Memory Fallacy](https://www.getzep.com/blog/the-ai-memory-wallet-fallacy/index.md): AI memory wallets sound appealing but face insurmountable economic, technical, and security challenges in practice.
- [Zep Q2 Feature Round-up](https://www.getzep.com/blog/zep-q2-feature-round-up/index.md): Zep Q2: Custom entities/edges, episode search, web debug/API logs, and more flexible, observable knowledge graphs for your AI apps.
- [Stop Using RAG for Agent Memory](https://www.getzep.com/blog/stop-using-rag-for-agent-memory/index.md): Here's why you shouldn't be using RAG for agent memory—and what to do instead.
- [Introducing Entity Types: Smarter, Structured Memory for Agents](https://www.getzep.com/blog/entity-types-structured-agent-memory/index.md): Zep's new Entity Types let developers precisely structure and recall domain-specific information for more accurate, personalized agents.
- [Lies, Damn Lies, & Statistics: Is Mem0 Really SOTA in Agent Memory?](https://www.getzep.com/blog/lies-damn-lies-statistics-is-mem0-really-sota-in-agent-memory/index.md): Mem0 claims State-of-the-Art in Agent Memory, but Zep outperforms it by 24%. We unpack why.
- [GPT-4.1 and o4-mini: Is OpenAI Overselling Long-Context?](https://www.getzep.com/blog/gpt-4-1-and-o4-mini-is-openai-overselling-long-context/index.md): We put OpenAI’s latest models through the LongMemEval benchmark—here’s why raw context size alone isn't enough.
- [The One-Token Trick](https://www.getzep.com/blog/the-one-token-trick/index.md): How single-token LLM requests can improve RAG search at minimal cost and latency.
- [Announcing a New Direction for Zep's Open Source Strategy](https://www.getzep.com/blog/announcing-a-new-direction-for-zeps-open-source-strategy/index.md): Ending support for Zep Community Edition to fully focus our open-source efforts on Graphiti
- [Introducing Metered Billing and BYOC Deployments](https://www.getzep.com/blog/introducing-metered-billing-and-byoc-deployments/index.md): Flexible, Scalable Pricing for High-Growth Companies and Enhanced Data Privacy for Enterprises
- [Cursor IDE: Adding Memory With Graphiti MCP 🤖⚡️](https://www.getzep.com/blog/cursor-adding-memory-with-graphiti-mcp/index.md): Cursor Agent can now remember your preferences, standards, and specs across sessions and projects.
- [Zep Q1 Product Round-up](https://www.getzep.com/blog/zep-q1-2025-product-round-up/index.md): Graph Explorer, an upgraded Playground, and a practical Cookbook—giving developers more control, better graph quality, and improved usability.
- [Building a Memory Agent with the OpenAI Agents SDK and Zep](https://www.getzep.com/blog/building-a-memory-agent-with-the-openai-agents-sdk-and-zep/index.md): A video walkthrough demonstrating using Zep's agent memory and the new OpenAI Agents SDK to build an AI agent with long-term memory.
- [Zep Feature Retirements: May 2025](https://www.getzep.com/blog/zep-feature-retirements-may-2025/index.md): Several features will be retired after May 31, 2025. Read this post for important information on migrating to alternatives.
- [🎉 Big News! Graphiti is Launching on Product Hunt](https://www.getzep.com/blog/big-news-graphiti-is-launching-on-product-hunt/index.md): Graphiti to launch on Product Hunt this Wednesday, February 19th!
- [Zep Is The New State of the Art In Agent Memory](https://www.getzep.com/blog/state-of-the-art-agent-memory/index.md): Setting a new standard for agent memory with up to 100% accuracy gains and 90% lower latency.
- [Zep: A Temporal Knowledge Graph Architecture for Agent Memory](https://www.getzep.com/blog/zep-a-temporal-knowledge-graph-architecture-for-agent-memory/index.md): We introduce Zep, a novel memory layer service for AI agents that outperforms the current state-of-the-art systems.
- [December 2024 Round-up](https://www.getzep.com/blog/knowledge-graph-explorer/index.md): Explore User and Group Graphs, Fact Ratings, and SOC 2 Type 2 Certification!
- [How do you search a Knowledge Graph?](https://www.getzep.com/blog/how-do-you-search-a-knowledge-graph/index.md): Can you build graph search that's both elegant & powerful? Here's how we did it in Graphiti.
- [Building A Russian Election Interference Knowledge Graph](https://www.getzep.com/blog/building-a-russian-election-interference-knowledge-graph/index.md): How we built a Knowledge Graph-based app for exploring Russian interference in the run-up to the 2024 US elections.
- [Exploring Russian Election Interference with a Knowledge Graph](https://www.getzep.com/blog/russian-influence-operations-graph/index.md): We built a visual exploration tool and AI assistant for analyzing Russian election interference operations using our Knowledge Graph library, Graphiti.
- [Beyond Chat Memory: Making AI Interactions More Personal](https://www.getzep.com/blog/ai-knowledge-graph-memory/index.md): Zep now connects user conversations and business data to help AI agents understand and serve users better.
- [Beyond Static Graphs: Engineering Evolving Relationships](https://www.getzep.com/blog/beyond-static-knowledge-graphs/index.md): Knowledge Graphs aren't adept at modeling changes in facts.
- [Announcing: Zep Community Edition](https://www.getzep.com/blog/announcing-zep-community-edition/index.md): Zep Community Edition, open-sourced today, is the first Zep product powered by a Knowledge Graph, allowing you to build more accurate and personalized Agents.
- [Scaling LLM Data Extraction: Challenges, Design decisions, and Solutions](https://www.getzep.com/blog/llm-rag-knowledge-graphs-faster-and-more-dynamic/index.md): Making LLM Knowledge Graph building fast & dynamic: challenges, design decisions, and solutions
- [Graphiti: Temporal Knowledge Graphs for Agentic Apps](https://www.getzep.com/blog/graphiti-knowledge-graphs-for-agents/index.md): Graphiti builds dynamic, temporally aware knowledge graphs that represent complex, evolving relationships between entities over time.
- [Announcing: Zep Fact Ratings](https://www.getzep.com/blog/announcing-zep-fact-ratings/index.md): Zep's Memory for LLM Apps is powered by user Facts. Devs can significantly improve Fact relevance with Ratings.
- [Zep for Structured Outputs from Chat History (Video Walkthrough)](https://www.getzep.com/blog/llm-structured-data-extraction/index.md): An end-to-end walkthrough demonstrating how to quickly and accurately extract data from chat histories stored in Zep.
- [Launching Structured Outputs from Chat History](https://www.getzep.com/blog/struct