Memori

MemoryApache-2.0

Memori is a memory layer for LLM applications, agents and copilots. Developers register their LLM client with a Python or TypeScript SDK and tag calls with an entity and a process, so facts and preferences from earlier interactions are stored and recalled later. It works with several model providers and frameworks, and agents such as Claude Code can also connect over MCP without SDK work. It runs as the hosted Memori Cloud or against the developer's own database.

Memories are attributed with an entity identifier, a process identifier and a session; the README says memories cannot be made without attribution.

Pinout

  • MCP server: supported, verified
  • HTTP API: not checked yet
  • command line: supported, verified
  • Python SDK: supported, verified
  • TypeScript SDK: supported, verified
  • OpenAPI description: not checked yet
Where each pin comes from
InterfaceStatusSource we readRead on
MCP serverSupported, verifiedGitHub README, MemoriLabs/Memori, MCP
HTTP APINot checked yet——
Command lineSupported, verifiedGitHub README, MemoriLabs/Memori, CLI
Python SDKSupported, verifiedGitHub README, MemoriLabs/Memori, Install
TypeScript SDKSupported, verifiedGitHub README, MemoriLabs/Memori, Install
OpenAPI descriptionNot checked yet——

Facts

Facts about Memori
PricingBilled yearly: Team $60K a year for a single production agent, Business $150K a year for several production agents, Enterprise priced on request. Memori Labs pricing page, read .
PurposeA memory layer for LLM applications, agents and copilots. Memori Cloud docs, read .
ProvidersAnthropic, Bedrock, DeepSeek, Gemini, Grok and OpenAI, sync and async, streamed and unstreamed. GitHub README, MemoriLabs/Memori, read .
FrameworksAgno, LangChain and Pydantic AI. GitHub README, MemoriLabs/Memori, read .
MCPAgents such as Claude Code, Cursor, Codex, Warp and Antigravity connect over MCP with no SDK integration. GitHub README, MemoriLabs/Memori, MCP, read .
CLIpython -m memori manages account, keys and quotas. GitHub README, MemoriLabs/Memori, CLI, read .
HostingManaged cloud, single-tenant cloud, VPC and on-premises deployments are listed; bring-your-own-database setups are documented separately. GitHub README, MemoriLabs/Memori, read .
Sign-upAccounts are created at app.memorilabs.ai to get an API key. Memori Cloud docs, read .
LicenseApache-2.0. Repository LICENSE file, read .
Last commit. GitHub API, latest commit on the default branch of MemoriLabs/Memori, read .
FrameworksLangChain (GitHub README, MemoriLabs/Memori, read 2026-11-16); Pydantic AI (GitHub README, MemoriLabs/Memori, read 2026-11-16); Agno (GitHub README, MemoriLabs/Memori, read 2026-11-16).

Limits

  • The README states that without attribution Memori cannot make memories.
  • Only Python and TypeScript quickstarts are documented.

Strengths

  • SDKs in two languages plus an MCP connection are documented.
  • Managed cloud and bring-your-own-database setups are documented.
  • Six model providers are listed.

Pricing

PricingBilled yearly: Team $60K a year for a single production agent, Business $150K a year for several production agents, Enterprise priced on requestMemori Labs pricing page, read 16 Nov 2026

As written on Memori's own page on the read date. Prices change: check there before you pay.

Alternatives to Memori

All memory tools for agents
  1. Agent memory system with retain, recall and reflect operations.

    Memory
  2. Open-source memory layer storing agent context across sessions and models.

    Memory
  3. Memory API with a hosted MCP server, Python and TypeScript SDKs, and a local server.

    Memory
  4. Open-source memory platform that turns documents and conversations into searchable memory.

    Memory

Last check: . Method: public repository and docs read automatically, page and day written down. Next check due within 90 days. How we check

Is this your tool? Write to hello@agent-tools.directory from an address on its domain to correct a pin or a fact.