Redis Agent Memory

What it does

Redis Agent Memory is a managed memory service for AI applications. It stores ordered conversation events (with actor, role, content, timestamps, and metadata) under a session ID as session memory, and keeps durable information such as preferences or facts in long-term memory. When summarization is enabled, older session events are condensed while recent events stay in full.

Developers reach it through the Python and TypeScript SDKs or a REST API, and it works with any agent framework or LLM provider. Long-term memory can be searched with semantic, keyword, or hybrid search and filtered by owner, session, namespace, topic, or memory type. Custom memory types define structured fields and extraction instructions, and sensitive-data exclusions steer extraction away from information that should not be kept.

It suits teams building copilots, chatbots, internal assistants, or agentic workflows that need context to persist across conversations. Setup is through Redis Cloud, or a private preview for Redis Software on Kubernetes, and the product is documented as preview with behavior subject to change.

What it can do

• Store session-scoped conversation events with role, timestamps, and metadata, retrievable by session ID
• Automatically summarize older session events while keeping recent ones in full
• Asynchronously extract durable long-term memories from session events, with vector embeddings
• Search long-term memory via semantic, keyword, or hybrid search with metadata filters
• Define custom memory types with structured fields and extraction instructions
• Exclude specified sensitive information from automatic extraction

Works with

• Python SDK
• TypeScript SDK
• REST API

Before you choose

• Documented as a preview; features and behavior are subject to change.
• Agent Memory billing is not stated on the Redis pricing page, so confirm costs with Redis before production use.
• Session memory and long-term memory each have their own configurable retention (TTL).

Listed under

• Coding
• MCP and agent tools

Get started

Create a managed Redis Agent Memory service, then follow the Python, TypeScript, or REST quickstart.

1. Create a Redis Cloud managed Redis Agent Memory service (or join the Redis Software private preview).
2. Choose a client: Python SDK, TypeScript SDK, or REST API.
3. Follow the quickstart to store session events, enable extraction and summarization, and retrieve long-term memories.

Redis Cloud database tiers are Free ($0, up to 30 MB), Essentials (from $0.007/hour) and Pro (from $0.014/hour, minimum $200/month). How Agent Memory is billed, or whether it is covered by these tiers, is not stated, so confirm with Redis before purchase.

We couldn’t find published prices for Redis Agent Memory. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Redis Agent Memory | Docs ↗redis.io
• Redis Agent Memory | Redis ↗redis.io
• Pricing | Redis ↗redis.io

Company website

redis.io: product, pricing and policy pages

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.