What is Zep?
Zep ingests chat history, JSON, and documents into a temporal knowledge graph that tracks how facts about entities change over time, then assembles ranked context for an LLM in under 200ms. Sold to engineering teams building personalized agents who find vanilla RAG too lossy and stateless. The temporal-graph approach (vs flat vector memory) is the differentiator.
Tools for building, hosting, testing, observing, connecting, and giving memory or computer access to AI agents.
See the full Agent Infrastructure guide to compare more tools, buyer criteria, and related workflows.
Use cases to evaluate
Personalized assistant that tracks evolving user facts over months
Sales agent merging CRM data with chat history per account
Customer-support agent that knows a ticket's full history and resolution
Healthcare agent under HIPAA BAA combining EHR snippets and conversation
Fit to evaluate
Engineering leaders shipping personalization without a dedicated ML team
Devs frustrated with vector-only RAG for stateful agents
B2B SaaS adding memory to multi-user copilots
Regulated industries needing SOC 2 Type II / HIPAA-compliant memory
Business fit
Right for you if your agent needs to know that 'Alice's address changed last month' rather than retrieving both the old and new address with equal weight. Useful when combining structured business data with chat. Skip if your use case is straightforward semantic search over docs, or if the credit-based pricing (episodes by byte size) is hard to forecast at your volume.
How to evaluate Zep
Use this category when a business wants agents that do work across tools, APIs, browsers, and data sources.
Confirm the exact workflow
Map Zep to one concrete workflow first, such as personalized assistant that tracks evolving user facts over months. Avoid buying before the owner, trigger, output, and success metric are clear.
Check category fit
Compare tool-calling, memory, browser automation, evals, observability, and deployment controls.
Compare practical alternatives
Shortlist Zep against Orgo, Browser Use, Browserbase so the decision is based on fit, effort, and workflow ownership rather than brand recognition alone.
Validate cost and rollout effort
Free: 1K credits/mo (no rollover), 2 projects. Flex $125/mo: 50K credits then $25 per 10K, 600 req/min, 5 projects, 1-day logs. Flex Plus $375/mo: 200K credits then $75 per 40K, 1K req/min, 10 projects, 7-day logs, webhooks. Enterprise custom (SOC 2 Type II, HIPAA BAA, BYOK/BYOM/BYOC). Credits: 1 per 350 bytes of episode; webhook = 1/8 credit. Also confirm implementation time, support needs, and whether the technical setup matches your team.
Compare Zep with alternatives
Use this quick comparison before booking demos or moving data into a new system.
| Primary workflow | Personalized assistant that tracks evolving user facts over months, Sales agent merging CRM data with chat history per account |
|---|---|
| Best-fit team | Engineering leaders shipping personalization without a dedicated ML team, Devs frustrated with vector-only RAG for stateful agents |
| Implementation effort | Technical setup and maintenance profile |
| Pricing check | Free plan + paid plans |
| Closest alternatives | OrgoBrowser UseBrowserbaseHyperbrowser |
Zep pricing
| Model | Free plan + paid plans |
|---|---|
| Snapshot | Free: 1K credits/mo (no rollover), 2 projects. Flex $125/mo: 50K credits then $25 per 10K, 600 req/min, 5 projects, 1-day logs. Flex Plus $375/mo: 200K credits then $75 per 40K, 1K req/min, 10 projects, 7-day logs, webhooks. Enterprise custom (SOC 2 Type II, HIPAA BAA, BYOK/BYOM/BYOC). Credits: 1 per 350 bytes of episode; webhook = 1/8 credit. |
| Checked |
Common questions about Zep
What is Zep?
Zep ingests chat history, JSON, and documents into a temporal knowledge graph that tracks how facts about entities change over time, then assembles ranked context for an LLM in under 200ms. Sold to engineering teams building personalized agents who find vanilla RAG too lossy and stateless. The temporal-graph approach (vs flat vector memory) is the differentiator.
What is Zep used for?
Common use cases: Personalized assistant that tracks evolving user facts over months; Sales agent merging CRM data with chat history per account; Customer-support agent that knows a ticket's full history and resolution; Healthcare agent under HIPAA BAA combining EHR snippets and conversation.
How much does Zep cost?
Free: 1K credits/mo (no rollover), 2 projects. Flex $125/mo: 50K credits then $25 per 10K, 600 req/min, 5 projects, 1-day logs. Flex Plus $375/mo: 200K credits then $75 per 40K, 1K req/min, 10 projects, 7-day logs, webhooks. Enterprise custom (SOC 2 Type II, HIPAA BAA, BYOK/BYOM/BYOC). Credits: 1 per 350 bytes of episode; webhook = 1/8 credit.
Who is Zep best for?
Zep fits Engineering leaders shipping personalization without a dedicated ML team, Devs frustrated with vector-only RAG for stateful agents, B2B SaaS adding memory to multi-user copilots, Regulated industries needing SOC 2 Type II / HIPAA-compliant memory. Right for you if your agent needs to know that 'Alice's address changed last month' rather than retrieving both the old and new address with equal weight. Useful when combining structured business data with chat. Skip if your use case is straightforward semantic search over docs, or if the credit-based pricing (episodes by byte size) is hard to forecast at your volume.
What are alternatives to Zep?
Common alternatives to Zep include Orgo, Browser Use, Browserbase, Hyperbrowser, Steel, Anchor Browser.