Memable
Memable is an AI memory platform that extracts knowledge from your sources into persistent semantic memory for AI agents.
Last verified:
What is Memable?
Memable is an AI-native bookmark and recall engine that extracts knowledge from your meetings, notes, tickets, and docs into persistent AI memory. Unlike traditional RAG systems that store document chunks and return similar text, Memable extracts actual facts, decisions, preferences, and procedures, storing them as actionable memories that learn and consolidate over time. The tool connects to URLs (live), files (PDFs, docs, text), S3/R2 cloud storage, MCP-compatible tools like Claude and Cursor, and soon Zendesk, Notion, and Slack.
Key features include private memory spaces for personal use, shared memory spaces for teams, automatic memory capture from coding sessions via Claude Code, an MCP server for seamless tool integration, and a recall API that returns actionable answers instead of document lists. The platform offers up to 1,000 memories free, unlimited memories on Pro, 3 team seats included, and priority support. Memable links memories back to their original sources while being 100x more efficient than RAG (100 memories vs 10,000 chunks).
Memable is designed for solo developers, freelancers, power users, and teams in support, sales, or any department that shares context. It solves the problem of AI forgetting everything between sessions and team knowledge scattered across 47+ tools like Zendesk, Slack, and Notion. The tool works with Claude Desktop, Cursor, and any MCP-compatible tool through simple 3-file setup.
Memable pricing
Pricing model: Freemium
$0/month for Free tier: Up to 1,000 memories, 3 MCP profiles, all connectors, personal spaces. Pro tier: Unlimited memories, unlimited MCP profiles, team spaces & invites (3 seats included), all connectors, priority support. Free trial available for Pro. Free to start with no credit card required. Team pricing includes 3 seats included in Pro plan.
Memable pros
- Extracts actual knowledge facts instead of document chunks
- Returns actionable memories instead of similar text
- Learns and consolidates memories over time
- 100x more efficient than RAG (100 memories vs 10k chunks)
- Links memories back to original source URLs
- Automatic memory capture from Claude Code coding sessions
- Works with Claude Desktop, Cursor, and any MCP tool
- Simple 3-file setup for Claude Code integration
- Up to 1,000 memories on free tier
- Unlimited memories on Pro plan
- 3 team seats included in Pro
- Shared memory spaces for team collaboration
- Private memory spaces for personal use
- All connectors available on free tier
- Priority support on Pro plan
- Live URL extraction for instant knowledge capture
- Cloud storage sync via S3 and R2
- No credit card required to start
- Free trial available for Pro
- Admin API for embedding in other platforms
Memable cons
- Zendesk, Notion, and Slack connectors not yet available (marked 'Soon')
- Team seats limited to 3 on Pro plan
- Only 1,000 memories on free tier (unlimited requires Pro)
- Only 3 MCP profiles on free tier
- Stop hook requires manual API key setup in shell profile
- Stop hook doesn't inherit env vars from claude.json
- Primarily focused on developer/AI tool workflow
- No mobile app mentioned on website
Frequently asked questions about Memable
What is Memable and how does it work?
Memable is an AI-native bookmark and recall engine that extracts knowledge from your sources (URLs, files, S3 buckets, MCP tools) and turns it into AI memory. It connects to your knowledge sources, uses AI-powered extraction to find facts, decisions, preferences, and procedures, then makes this memory accessible via MCP to your AI tools or directly through the API.
How is Memable different from RAG?
RAG stores document chunks and returns similar text statically, never learning, and needs huge context windows. Memable stores extracted knowledge, returns actionable memories, learns and consolidates over time, and is 100x more efficient (100 memories vs 10,000 chunks). RAG gives you 5 similar tickets to read; Memable gives you the instant answer: 'This error means X, fix by doing Y.'
What connectors does Memable support?
Live connectors include: URLs (paste a link for instant extraction), Files (PDFs, docs, text), S3/R2 (cloud storage sync), MCP (Claude, Cursor, any tool), and Claude Code (auto-capture from coding sessions). Soon connectors include: Zendesk (support ticket knowledge), Notion (wiki and docs), and Slack (team conversations). More connectors are added every month.
How do I set up Memable with Claude Code?
Set up in 3 files: 1) ~/.claude.json with MCP server config including MEMABLE_API_KEY and MEMABLE_API_URL, 2) ~/.claude/settings.json with Stop hook that runs 'npx memable@latest extract-session', 3) ~/.claude/CLAUDE.md noting to call 'boot' at session start and 'recall' before architectural decisions. Add API key to shell profile (~/.zshrc or ~/.bashrc) since Stop hook doesn't inherit env vars.
What is the difference between Personal and Teams usage?
Personal provides private memory spaces just for you, where your AI knows YOUR preferences—ideal for solo devs, freelancers, and personal use with up to 1,000 memories free. Teams provides shared memory spaces for your whole team so everyone stays in sync, with 3 seats included, perfect for support, sales, or any team that shares context, with unlimited memories.
What pricing plans does Memable offer?
Free: $0/month with up to 1,000 memories, 3 MCP profiles, all connectors, and personal spaces. Pro: Unlimited memories, unlimited MCP profiles, team spaces & invites (3 seats included), all connectors, and priority support. Free trial available for Pro. No credit card required to start.
Which AI tools work with Memable?
Memable works with Claude Desktop, Cursor, and any MCP-compatible tool. You add it as an MCP server through simple JSON config (claude_desktop_config.json or ~/.claude.json) with command 'npx memable' and your API key. Once configured, your AI tool has instant access to all your memories.
Can I embed Memable in my own platform?
Yes, Memable offers an Admin API for building platforms that want to embed Memable functionality. The website states 'Building a platform? Embed memable with our Admin API' with a link to learn more about the embedding options.
How does Memable prevent me from forgetting important information?
Memable extracts knowledge from all your sources into persistent AI memory that doesn't forget between sessions. Your AI learns from everything you know, storing facts, decisions, preferences, and procedures as actionable memories. The system links memories back to sources and provides efficient recall (100 memories vs 10,000 chunks), so your AI actually knows your preferences and can give instant actionable answers.
Do I need a credit card to start using Memable?
No, Memable is free to start with no credit card required. You can use the Free tier with up to 1,000 memories, 3 MCP profiles, and all connectors. You only upgrade to Pro when you need unlimited memories and team features.