Supermemory

Agent-enablingFreemium

Context and persistent memory infrastructure for AI agents

Overview

About Supermemory

This summary uses submitted and public product material. Only source-mapped claims are documentation reviewed.

Supermemory provides an MIT-licensed memory and context layer for agents, combining extracted memories, evolving user profiles, document retrieval, connectors, and hosted or self-hosted operation.

So agents can…

Retain facts and user context across sessions
Retrieve documents and memories through one context layer
Update profiles as information changes over time

Listing decision

Why Supermemory is listed

Supermemory supplies memory ingestion, profile building, retrieval, search, and connectors to other agents and applications rather than deciding their goals or actions, so it is agent-enabling.

Before you choose

Fit and limitations

Best for

Agents and AI products that need durable user memory, document retrieval, profiles, context search, connectors, or a local context engine without assembling separate embedding, chunking, and vector-database components.

Known limitations

Hosted usage draws from plan balances, the free plan pauses when its balance is exhausted, and production self-hosting is offered on higher tiers. Connector access varies by plan, while local or air-gapped operation still requires operators to choose, host, and secure the model and surrounding infrastructure.

Not yet verified

Performance and benchmark leadership claims in first-party pages were not independently reproduced. This review also did not establish feature parity among the hosted API, open-source/local runtime, and commercial self-hosted deployments.

Sources

Sources and verification

First-party unless marked otherwise. Each link supports only its adjacent claim.

Supermemory quickstart — memory API and SDK access
The official quickstart documents authenticated memory add-and-retrieve calls through the Supermemory API and its TypeScript and Python SDKs.
official documentation · accessed Sep 3, 2026
Supermemory repository — memory, profiles, search, and local operation
The maintained repository describes a memory and context engine with fact extraction, user profiles, hybrid search, connectors, multimodal processing, agent plugins, MCP access, and a local runtime that can use Ollama offline.
official repository · accessed Sep 3, 2026
Supermemory pricing — hosted usage and deployment options
The pricing page lists a free API plan and paid Pro, Max, Scale, and Enterprise tiers, usage-based context primitives, self-hosting on Scale and Enterprise, and air-gapped deployment on Enterprise.
official pricing · accessed Sep 3, 2026
Supermemory repository license — MIT
The maintained Supermemory repository is published under the MIT License.
official license · accessed Sep 3, 2026