Skip to main content
AgentsUse

Category

Agent Memory

Memory layers that let agents remember users, facts, and prior work across sessions.

How to read this category

Context windows forget. Memory layers are what an agent keeps between sessions: facts about the user, outcomes of past work, and how those facts changed.

Mem0 extracts and stores facts with fast recall, open source or hosted, and is the quicker add to an existing agent. Zep builds memory on a temporal knowledge graph, so the agent can know what was true at a given time rather than only the latest version of a fact.

Memory is also a liability. Stored facts go stale, bleed across users, and get stated with confidence. Whichever you pick, source stamps and expiry are your job to configure, not features to assume.

The verified profiles

Open source, Apache-2.0Mem0Memory layer for AI agents that persists user, session, and agent context across conversations for personalized applications. Best for: Developers adding persistent, personalized memory to assistants, support bots, and autonomous agents with a simple add and search API. By Mem0.Freemium (hosted paid tiers)ZepManaged agent memory platform that builds temporal context graphs from conversations and business data for governed retrieval. Best for: Enterprise teams that need temporal, graph-based memory over changing conversations and business data with governance, policy filtering, and managed or VPC deployment. By Zep.

Side by side on sourced facts

ToolPricing modelLicenseDeploymentMCP supportChecked
Mem0Freemium (hosted paid tiers)Apache-2.0library, self-hosted, cloudMCP server option2026-10-07
ZepFreemium (hosted paid tiers)Not open sourcecloud, cloud-byok, byocRemote Memory MCP Server2026-10-07

Every cell traces to the tool profile, which traces to the official repository, package page, and documentation checked on the date shown. Pricing cells name the model, not a price: vendor prices change, so check the profile for the sourced pricing summary and the official pricing link.

What actually decides the choice

  • Decide if you need history or just recall. Timeline questions (what changed, when) point at a graph; simple per-user facts do not.
  • Open-source and self-hosted paths exist here; weigh them against the operational cost of another stateful service.
  • Write policy beats read speed: what gets stored, for whom, and when it expires matter more than recall benchmarks.

How this archive is ordered, and why it is short

Profiles appear in the order they passed the AgentsUse review gate, not in a tested quality ranking. AgentsUse has not run head-to-head benchmarks for this category, so this page does not claim a winner. The comparison table is the ranking tool: sort by the dimension your job cannot compromise on, then read the full profiles for limitations.

This archive currently lists 2 verified profiles. AgentsUse normally asks for eight published profiles before opening a category archive. This one opens early, on the record, because every listed profile passed the full tool gate and readers comparing real choices need the side-by-side surface now, not a longer unvetted list later. The archive grows as more tools pass review. See how verification works.

Read before you give these tools real work

Guide

Managing Agent Memory and Context

Agents forget, drift, and drown in their own context. How memory actually works, what fills it fastest, and patterns that keep long runs coherent.

2026-09-29

Guide

Building Agent Error Recovery

Agents fail; the question is whether failure is graceful or catastrophic. A taxonomy of agent failures and the recovery patterns - retries, fallbacks, circuit breakers - that contain them.

2026-09-29

Back to the tools directory, the category index, or the guides.