kapa.ai vs Zep (2026)

kapa.ai

by kapa.ai, Inc.

Zep

by Zep AI

Pros

• ✓Every generated answer carries inline citations to the customer's own sources, and it is designed to say it does not know when those sources do not cover a question.
• ✓One knowledge base serves the Retrieval API, a hosted MCP server and pre-built agents for a website widget, Slack, Discord, a support form, Zendesk and in-product use.
• ✓Ingests more than 20 source types, from documentation sites, GitHub and OpenAPI specs to Zendesk tickets, Confluence and Notion, with continuous sync on paid plans.

Limitations

• –Growth and Enterprise are quote-only, commercial plans carry a minimum annual commitment, and there is no self-serve card signup.
• –Private sources, continuous sync, zero data retention and PII masking start on Growth, and SSO, SCIM, role-based access control and EU hosting are Enterprise features.
• –The hosted MCP search tool returns results in about 3 seconds at the median and 4.5 seconds at the 95th percentile, and public OAuth servers allow 600 searches per user per day.

Pros

• ✓Temporal knowledge graph automatically tracks how facts change over time. When information is updated or contradicted, Zep invalidates outdated facts while preserving historical context, making it more accurate than vector-only memory systems for complex multi-session agent workflows.
• ✓Zep publishes retrieval accuracy of 94.7% at 155ms on LoCoMo and 90.2% at 162ms on LongMemEval, its own measurements on two industry benchmarks. Single-shot retrieval avoids the slower agentic tool-call loops used by some memory systems.
• ✓Graphiti MCP server gives Claude Desktop, Cursor, and other MCP clients persistent knowledge graph memory across sessions. Python, TypeScript, and Go SDKs with webhook support and no agent framework lock-in, backed by named enterprise customers including Samsung, Zscaler and HoneyBook.

Limitations

• –Steep pricing jump from free tier (10,000 credits per month for prototyping) to Flex at $125 per month ($104 annual) with no intermediate tier. Early-stage teams that outgrow free credit limits face a significant cost increase with no middle-ground option.
• –Credit-based pricing makes costs difficult to forecast at production scale. Episodes consume 1 credit per 350 bytes, and paid plans auto-purchase more credits whenever the balance falls below 20% - Flex adds 10,000 credits at $25, Flex Plus 40,000 at $75 - so monthly spend is not capped.
• –Self-hosted Graphiti requires Docker, FalkorDB or Neo4j, and LLM API key management. The managed Memory MCP Server removes that setup but is metered in per-account MCP seats that vary by plan, and custom OIDC and UserGroup ABAC require the Enterprise plan.