Comparisons

claude-mem alternatives for coding agents.

Looking for a claude-mem alternative? Eight options, from the memory built into Claude Code to open-source MCP memory servers and memory a whole team's agents share: what each is for, how it connects and what it costs.

  • 8 alternatives, each linked to its source
  • Checked October 15, 2026
  • The best pick by need, then the full table
  • Session memory and engineering memory, told apart
How far each memory reachesas of October 2026
  1. Every teammate, every agentRecallium
  2. Every teammate, through gitA project CLAUDE.md
  3. One developer, several agentsagentmemory, Hindsight
  4. One developer, one machineclaude-mem, auto memory
  5. One sessionThe chat history

claude-mem records far more of each session than auto memory, but both stay with one developer.

The best claude-mem alternative, by what you need

It depends on why you are switching. As of October 2026:

  • Less to install, still in Claude Code: Claude Code’s built-in memory: CLAUDE.md you write, plus auto memory Claude keeps itself
  • Open source and local, across several agents: agentmemory (Apache-2.0), Basic Memory (AGPL-3.0, Markdown you can read) or Hindsight (MIT)
  • A memory service you also use in your own product: Mem0 or Supermemory, both with Claude Code plugins
  • A team’s agents sharing decisions and fixes across Claude Code, Codex, Cursor and Copilot: Recallium (this site): why it is built for that

Session memory or engineering memory?

Session memory

For one developer’s agent, on one machine

Memory that records what your agent did in each session and brings the relevant parts back next time. claude-mem does this automatically for Claude Code; Claude Code’s own auto memory keeps a lighter set of notes; agentmemory and Hindsight capture sessions across several agents.

Engineering memory

For the team and every coding agent it uses

The team’s record of why the code is the way it is: the decisions, root causes, fixes and rules its engineers and agents work out, linked to files and commits, and read by every teammate’s agent in any tool. Recallium is built for this.

claude-mem and 8 alternatives, compared

OptionWhat it is and who it’s forHow it connects to coding agentsDeployment and pricing
RecalliumPricingWhy RecalliumWhat it isEngineering memory for engineering teams: what the team works out in agent sessions, from why it was built this way to what broke and what fixed it, linked to files.How it connectsOver MCP: Claude Code, Codex, Cursor, GitHub Copilot and 60+ MCP clients; one installer wires 16+ coding agents.Deployment and pricingManaged cloud, priced per engineer; Cloud Free has 500 memories a month until December 2026. Join the waitlist.
claude-memGitHubCMEMWhat it isA memory plugin built for Claude Code: it captures what the agent does in each session, compresses it with an AI model and brings relevant context back in later sessions, on your machine.How it connectsLifecycle hooks plus MCP search tools, with a local web viewer. Claude Code first; also Codex, Gemini, Copilot, OpenCode and more.Deployment and pricingApache-2.0 open source with a local worker, SQLite and Chroma. Compression spends model tokens; CMEM Pro ($30/mo) adds hosted observation and cloud sync.
Claude Code memory (CLAUDE.md and auto memory)Memory docsWhat it isBuilt into Claude Code: CLAUDE.md files you write (project, user or org scope) and auto memory, notes Claude writes itself from your corrections and preferences.How it connectsLoaded at the start of every Claude Code session. A project CLAUDE.md is shared with the team through version control; auto memory is per repository and machine-local, “not shared across machines or cloud environments”.Deployment and pricingPlain files in the repository and on your machine; part of Claude Code, with no separate product to buy.
agentmemoryGitHubWhat it isPersistent memory for coding agents that captures sessions through hooks and finds them again with keyword, vector and graph search; shared and private namespaces.How it connectsAdapters for Claude Code, Copilot CLI, Cursor, Codex, Gemini and others, plus any MCP client.Deployment and pricingApache-2.0 open source, local-first on SQLite with no external services; free.
Basic MemoryHomePricingWhat it isKnowledge kept as plain Markdown files you own, with a full-text, semantic and knowledge-graph index on top; for individuals and teams who want memory they can read.How it connectsOver MCP; works with Claude, Codex, Cursor and VS Code, and with Obsidian for the files.Deployment and pricingFree AGPL-3.0 open source that runs locally; cloud sync with shared workspaces from $15/seat/mo (Team) or $30/seat/mo (Business).
HindsightGitHubCloud billingWhat it isAgent memory from Vectorize that retains, recalls and reflects on what an agent learns. It can also build per-repository memory for coding agents from git history.How it connectsEvery server has a built-in MCP endpoint; it installs repository memory for Claude Code, Codex CLI and Cursor CLI.Deployment and pricingMIT-licensed open source you run with Docker, pip or Helm, or Hindsight Cloud with prepaid usage.
Mem0PricingClaude Code pluginMCPWhat it isA memory layer you build into an AI application so it remembers its users: facts and preferences extracted from conversations, scoped by user, agent or run. “Unlimited end users” on every tier.How it connectsPython and JavaScript SDKs, a hosted MCP server, and plugins for Claude Code, Cursor, Codex, Windsurf and OpenCode; the Claude Code plugin shares one namespace per repository.Deployment and pricingManaged platform, or the Apache-2.0 open-source library. Metered by add and retrieval requests: Hobby free, Starter $19/mo, Pro $249/mo, Enterprise custom.
SupermemoryPricingClaude Code pluginMCPWhat it isA memory and context engine you build into an AI application so it remembers its users: documents, connector content and conversation memories. Pro is “3 team seats, unlimited end users”.How it connectsSDKs, a hosted MCP server, and Claude Code and Cursor plugins; the Claude Code plugin keeps project memory in a per-repository container, separate from personal memory.Deployment and pricingManaged service, plus MIT-licensed open source that can run locally; self-hosted deployments on Scale and Enterprise. Credits for tokens, queries and operations: Free, Pro $19/mo, Max $100/mo, Scale $399/mo.
MCP reference memory serverGitHubWhat it isA minimal knowledge-graph memory from the Model Context Protocol project: entities, relations and observations kept in one local file.How it connectsAny MCP client, including Claude Code, Codex and Cursor.Deployment and pricingMIT-licensed reference server you run locally; free.

Checked October 15, 2026against each vendor’s own documentation and pricing pages, linked under each name. Features, prices and availability change, so follow the links for current terms. All nine tools compared, with when to use what.

Recallium vs claude-mem, feature by feature

FeatureRecalliumclaude-mem (thedotmack/claude-mem)
Built forRecalliumEngineering teams and their coding agentsclaude-memClaude Code users who want sessions remembered automatically; also Codex, Gemini, Copilot and OpenCode
What it capturesRecalliumWhat the team works out in agent sessions: why it was built this way, what was decided, what broke and what fixed it, linked to filesclaude-memWhat the agent does in each session, compressed by an AI model into observations and injected into later sessions
Memory scopeRecalliumA project per repository, organized into workstreams; shared by every teammate’s agent on Cloud Proclaude-memYour machine in the open-source engine; a team pilot (TeamBrain) is listed with pricing not yet announced
DeploymentRecalliumManaged cloudclaude-memA local worker with SQLite and Chroma, installed with Bun and uv; optional hosted observation and cloud sync
Connects toRecalliumClaude Code, Codex, Cursor, GitHub Copilot and 60+ MCP clients; one installer wires 16+ coding agentsclaude-memLifecycle hooks plus MCP search tools and a local web viewer
PricingRecalliumPriced per engineer; Cloud Free has 500 memories a month until December 2026claude-memFree, Apache-2.0; compression spends model tokens; CMEM Pro $30/mo
LongMemEval-SRecallium99.8% hit@10 · 98.4% QA, full protocol publishedclaude-memNo LongMemEval-S result in its README or docs

Product details checked October 15, 2026, against the claude-mem repository, documentation and cmem.ai; claude-mem ships often, so follow the links for current terms. Its license is Apache-2.0 today; older write-ups say AGPL-3.0.

How Recallium and claude-mem differ in practice

Recallium

Engineering memory for the engineers and coding agents who build it

Recallium keeps the team’s reasoning rather than a log of sessions: the approach and the decision, the root cause and the fix, each linked to its files and commit, with workstreams to hand work between sessions and teammates and team rules loaded into every session. Every teammate’s agent reads the same memory, in Claude Code, Codex, Cursor or Copilot.

claude-mem

Automatic session memory for Claude Code, on your machine

claude-mem records what your agent does without you asking: it captures each session through hooks, compresses it with an AI model and puts the relevant parts back into the next session, with a local viewer to browse it. It is open source, free to run, and has one of the largest communities of any coding-agent memory tool.

Both connect over MCP, and both work with Claude Code and Codex.

When to choose Recallium

  • More than one engineer works in the repository and their agents should share what each one works out
  • Your team uses several agents, such as Claude Code with Cursor or Codex, on the same code
  • You want decisions, root causes and fixes recorded with who, which files and which commit
  • You want handoffs, team rules and commit-linked decisions, not only recall of past sessions
  • You would rather not run a background worker on every machine

When to choose claude-mem

  • You work alone and want every session captured automatically
  • You want everything to stay on your own machine, free and open source
  • You want to browse your session history in a local viewer

Questions

What is claude-mem?

An open-source memory plugin for Claude Code. It captures what the agent does during sessions, compresses it with an AI model and injects relevant context into later sessions. It stores memory locally in SQLite and Chroma and also supports Codex, Gemini, Copilot and OpenCode (GitHub, October 2026).

Why look for a claude-mem alternative?

The reasons raised most in its issue tracker are the model tokens spent compressing sessions, the background worker’s setup and resource use, and files it writes into the project; many of those issues are already closed. The other common reason is scope: the open-source engine remembers for one developer, and a team version is still a pilot.

Is claude-mem free?

The plugin is free and Apache-2.0 licensed. Compressing sessions uses a model, so it spends tokens from your Anthropic plan or a key you bring; CMEM Pro, at $30 a month, adds hosted observation and cloud sync (cmem.ai, October 2026).

claude-mem or Claude Code’s auto memory?

Auto memory is built in: Claude writes notes to MEMORY.md on your machine and loads the first 200 lines or 25KB at the start of each session. claude-mem captures far more of each session and lets the agent look it up. Neither is shared with teammates; see Claude Code memory.

Can a team share claude-mem?

Not with the open-source engine, which runs per machine. claude-mem lists a team pilot, TeamBrain, with pricing not yet announced. Recallium Cloud Pro shares one repository’s memory with every teammate’s agent.

Join the waitlist for Recallium Cloud, the managed engineering memory for your team.

Join the waitlist →