---
title: "claude-mem vs past.dev"
description: "claude-mem gives Claude Code persistent session context. past.dev provides temporal memory through an API. Compare their scope, storage and use cases."
canonical: https://past.dev/compare/claude-mem
last-updated: 2026-10-15
---
# claude-mem and past.dev

Source: https://past.dev/compare/claude-mem

claude-mem is a Claude Code plugin that records coding sessions, compresses them into observations and supplies relevant observations to later sessions. past.dev is a multi-user memory API for applications. It returns ranked documents with dates and source excerpts. The products can run together because they serve different scopes.

## What claude-mem does

[claude-mem](https://github.com/thedotmack/claude-mem) hooks into the Claude Code session lifecycle. It captures tool activity and outcomes, compresses them into typed observations and supplies relevant context to later sessions. It installs through the plugin marketplace and runs locally for one developer.

## What the memory API provides

`POST /api/v1/ingest` accepts timestamped text. Once ingestion completes, `POST /api/v1/recall` returns ranked documents with dates and source excerpts for a required identity. Project keys, audience slugs, and identity records control the public scoping contract. The [quickstart](/docs/memory-api/quickstart) shows the request flow.

## Scope comparison

| Question | claude-mem | past.dev |
| --- | --- | --- |
| Who is it for | A developer using Claude Code | A product and its users |
| Where memory lives | Locally, with the plugin | Behind an API, shared by design |
| What is sent | Session observations | Timestamped source text |
| A source changed | New observation accumulates | Stable source id replaces the prior revision |
| Unanswerable question | Whatever context surfaces | Application evaluates returned documents |
| Scope | Your sessions | Projects, audiences, and identities |

Use claude-mem for one developer's Claude Code sessions. Use an [agent memory](/what-is-agent-memory) API when an application must retain facts and history for its users.

## Running both

claude-mem can retain development-session context while application agents use past.dev. The [Claude Agent SDK integration](/integrations/claude-agent-sdk) shows the two-tool pattern. [Memory for OpenClaw](/integrations/openclaw) covers a similar setup for file-based agents. The [benchmarks page](/benchmarks) reports recall quality.

## Frequently asked questions

### Is past.dev a claude-mem alternative?

claude-mem is designed for developer session continuity in Claude Code. Use the past.dev API for product memory that requires timestamped input, identity scoping, and source-backed recall.

## Related

- [Memory for the Claude Agent SDK](https://past.dev/integrations/claude-agent-sdk)
- [Memory for OpenClaw](https://past.dev/integrations/openclaw)
- [What is agent memory?](https://past.dev/what-is-agent-memory)
- [Quickstart](https://past.dev/docs/memory-api/quickstart)