Memoryport
Memoryport is a local-first memory layer that gives Claude Code, Cursor, Ollama and OpenAI persistent context across every session.
Memoryport earns its place for developers juggling Claude Code, Cursor and Ollama who are tired of re-explaining context. Local LanceDB indexing plus a three-gate retrieval layer is a real engineering answer, not a prompt wrapper, and the offline guarantee is something cloud memory services can't match. The Free plan is genuinely usable — unlimited local storage, Claude Code and Cursor MCP servers, encryption at rest. Pro at $9.99/month buys Arweave backup, cross-device sync and cold-start rebuild from chain, and the memories survive after you stop paying, which is unusual for the category. The trade-off is CLI-only setup and desktop-only reach. If you live in a terminal and care about
Verified 1d ago · liveness 71/100 · cite: rightaichoice.com/tools/memoryport
- Developers using multiple AI coding assistants who want unified memory
- Privacy-conscious power users who want local-only AI memory
- Ollama and Open WebUI users wanting offline context retention
- Technical teams comfortable with CLI setup and self-hosting
- Non-technical users who need plug-and-play with no CLI setup
- Users who need mobile or web-based access to their memory store (desktop/CLI only)
- Teams requiring centrally managed cloud-hosted memory without self-hosting
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Memoryport if you need plug-and-play setup with no terminal, or you require mobile/web access or a centrally managed cloud tenant with admin controls.
Cross-device sync, Arweave backup and the cold start rebuild from chain are all Pro-only, so the $9.99/month kicks in the moment you want your memory on a second machine.
At $9.99/month Pro sits well below most cloud AI-memory subscriptions aimed at developers, and the Free plan is unusually complete for a local-only tier — unlimited local storage with Claude Code and Cursor MCP servers included. The paid step is really about backup and multi-device reach rather than unlocking core memory. If you need centrally managed cloud memory with admin controls, budget for enterprise vendors instead; Memoryport is deliberately not priced for that buyer.
In short
Memoryport — Memoryport is a local-first memory layer that gives Claude Code, Cursor, Ollama and OpenAI persistent context across every session. Best for Developers using multiple AI coding assistants who want unified memory, Privacy-conscious power users who want local-only AI memory, Ollama and Open WebUI users wanting offline context retention. Free to start; paid plans from $9.99/mo.
What people actually say about Memoryport — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
23 mentions across 3 sources (Hacker News, YouTube, GitHub) · researched Aug 30, 2026.
Average across the 3 sources that answered — each source counts once, not each post.
- +Local-first, permanent memory ensures privacy and data ownership
- +Integrates with Claude Code, Cursor, Ollama, and OpenAI via MCP
- +Open-source (Apache 2.0) with Rust codebase, allowing customization
- +Encrypted at rest with AES-256-GCM and Argon2id key derivation
- +Claims high performance with 500M token context and <300ms latency
- −Pro upgrade currently non-functional per GitHub issue
- −Limited community feedback; most YouTube comments are off-topic
- −CLI installation may be daunting for non-technical users
- −Benchmark claims not independently verified
- −No clear roadmap or release notes publicly shared
- • No hidden costs reported, but Pro tier may be non-functional, meaning you pay without getting service
Viability Score
How well maintained and how widely used is Memoryport? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Local-first memory store using LanceDB vector indexing
- One-command install via curl for macOS and Linux
- Interactive setup wizard (uc init) to configure integrations
- Three-gate retrieval system with query expansion
- MCP server for Claude Code
- MCP server for Cursor
- Ollama proxy integration
- Ollama native API proxy
- OpenAI/ChatGPT proxy integration
- Open WebUI support
- Support for any standard MCP client
- Knowledge graph visualization
- Session analytics dashboard
- AES-256-GCM encryption at rest with Argon2id key derivation
- Optional Arweave decentralized backup (Pro)
About Memoryport
Memoryport is an open-source, local-first memory layer that gives any LLM persistent context across sessions. You install it on macOS or Linux with a single curl command — `curl -fsSL https://memoryport.ai/install | sh` — then run the interactive `uc init` wizard to wire up your integrations. From then on every conversation is indexed locally in a LanceDB vector index, with nothing sent to the cloud unless you opt in. Context is injected automatically through a three-gate retrieval system with query expansion, so your AI gets relevant history rather than a dump of everything. It unifies memory across Claude Code, Cursor, Open WebUI, and Ollama or ChatGPT via MCP servers and API proxies, meaning one memory store serves all your AI tools. Data is encrypted at rest with AES-256-GCM and Argon2id key derivation. Memoryport publishes 97.9% session recall on LongMemEval (ICLR 2025), under 100ms p95 retrieval and 0.82 recall@10 across 500M tokens tested. The Free plan covers unlimited local storage; Pro at $9.99/month adds optional Arweave backup that survives disk failure, syncs across devices, and persists even if you stop paying. Built for developer and power-user workflows — CLI-first, Rust codebase, Apache 2.0 licensed.
Behind the Verdict
Most 'AI memory' products you can buy today ask you to hand your entire conversation history to someone else's server and trust that they encrypt it. Memoryport's pitch is the opposite: your machine is the server. The install is one curl command on macOS or Linux, followed by `uc init`, an interactive wizard that wires up whichever tools you use. After that you just use your AI normally — the vendor's own copy says 'nothing changes about your workflow' — and every conversation is indexed locally in LanceDB. What separates it from a weekend project is the retrieval layer. Memoryport uses a three-gate system with query expansion rather than stuffing prior context blindly, which matters because the failure mode of naive memory is that your model gets a wall of irrelevant history and gets worse, not better. The vendor publishes 97.9% session recall on LongMemEval (ICLR 2025), under 100ms p95 retrieval and 0.82 recall@10 at 500M tokens tested. Treat those as vendor-reported figures, but the architecture is at least aimed at the right problem. The multi-tool story is the other draw. MCP servers for Claude Code and Cursor, a proxy for Ollama (both native API and proxy modes), an OpenAI proxy for ChatGPT, and Open WebUI support mean one memory store follows you across editors and chat clients instead of each tool having its own amnesia. Anything speaking standard MCP should work too. Encryption at rest is AES-256-GCM with Argon2id key derivation, and even with cloud backup enabled the vendor states only you can read it. The codebase is Rust, Apache 2.0, and fully readable if you want to verify any of that. Where it gets awkward: setup assumes a terminal and a technical disposition. There's no plug-and-play installer, and it's desktop/CLI only — no mobile or web access to your memory. The Free plan keeps everything local; cross-device sync, Arweave backup, cold-start rebuild from chain and the ~250M token/month allowance sit behind the $9.99/month Pro plan. If your organization needs centrally managed cloud-hosted memory with admin controls and SSO, this is not that product and you should be looking at enterprise memory vendors instead. For the audience it's built for — developers running two or three AI assistants in parallel who are tired of re-establishing context — the Free plan is a low-risk way to test the retrieval quality on your own conversations, and Pro is a $9.99/month question you can answer after a week of real use.
Researching Memoryport? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Memoryport actually fits — and what changes day-one when you adopt it.
Install with the one-line curl command, run uc init to attach both MCP servers, then work across Claude Code for refactors and Cursor for UI work without restating the project brief.
Outcome: Both assistants pull from the same local LanceDB index, so context carries over between the two tools and across sessions without copy-pasting.
Point the Ollama proxy at Memoryport, keep the entire index on-machine, and use the knowledge graph visualization to see what the assistant has accumulated.
Outcome: Fully offline AI memory with AES-256-GCM encryption at rest and no cloud dependency, at $0 on the Free plan.
Upgrade to Pro at $9.99/month, enable Arweave backup, and later restore memory from chain onto a new laptop.
Outcome: Conversation history survives disk failure and moves across devices, with memories retained even if the Pro subscription lapses.
Use Cases
- Keep conversation history across LLM sessions so you never re-explain project context.
- Share one memory between Claude Code and Cursor on the same machine via MCP.
- Back up AI memory to Arweave for cross-device restore and disk-failure recovery.
- Run Memoryport with Ollama for fully offline AI memory with no cloud dependency.
- Give ChatGPT persistent cross-session context through the OpenAI proxy.
- Visualize knowledge graphs built from your accumulated LLM conversations.
- Rebuild memory from chain after a cold start or machine migration.
Limitations
- Memoryport skips the model layer — it doesn't ship an AI of its own, so it exists to feed context into Claude Code, Cursor, Ollama, OpenAI, and other MCP clients.
- Its local-first architecture means setup runs through a CLI curl command plus an interactive wizard, with one-command install supported on macOS and Linux.
- The Free plan is local-only: cross-device sync, Arweave backup, cold start rebuild from chain, and the ~250M token/month allowance all sit behind the $9.99/month Pro plan.
- If you need mobile or browser access to your memory, or a vendor-managed cloud tenant with centralized admin controls, this isn't that product.
as of 2026-10-10
Verification history
We have re-verified Memoryport 9 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
Showing the 6 most recent of 9 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Memoryport tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0
Ideal for
Solo developers and privacy-first power users running Claude Code, Cursor, Ollama or Open WebUI on one machine and happy to keep everything local.
What this tier adds
Starting tier — unlimited local storage, Claude Code and Cursor MCP servers, knowledge graph visualization, session analytics and encryption at rest at $0.
Pro
$9.99/month
Ideal for
Developers who work across more than one device or want their memory to survive a disk failure and remain readable even after they stop paying.
What this tier adds
Adds Arweave backup across hundreds of locations, cross-device sync, cold start rebuild from chain and a ~250M tokens per month allowance for $9.99/month.
Where the pricing makes sense
The company stage and team size where Memoryport's pricing actually pencils out — and where peers do it cheaper.
At $9.99/month Pro sits well below most cloud AI-memory subscriptions aimed at developers, and the Free plan is unusually complete for a local-only tier — unlimited local storage with Claude Code and Cursor MCP servers included. The paid step is really about backup and multi-device reach rather than unlocking core memory. If you need centrally managed cloud memory with admin controls, budget for enterprise vendors instead; Memoryport is deliberately not priced for that buyer.
Setup time & first value
How long it actually takes to get something useful out of Memoryport — broken out by persona, not the marketing-page minute.
Terminal-comfortable developers should reach first value in roughly 10–15 minutes: the curl install and uc init wizard handle macOS and Linux, and the wizard configures your MCP servers and proxies in one pass. Give it a little longer if you're attaching several clients (Claude Code, Cursor, Open WebUI, Ollama, OpenAI) or enabling the Pro Arweave backup, and expect indexing backlog if you import
Switching to or from Memoryport
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From manual context pasting: run the curl installer and uc init, then let the three-gate retrieval inject history automatically.
- →From per-tool chat history: attach Memoryport's MCP server to Claude Code and Cursor so one LanceDB index replaces each tool's separate memory.
- →From a cloud memory service: point the OpenAI and Ollama proxies at Memoryport and keep the index on your own machine instead.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Memoryport”, and we withheld 6: 6 could not be judged, because “Memoryport” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Memoryport.
Official links
Tools that pair well with Memoryport
Common stack mates teams adopt alongside Memoryport, with the specific reason each pairing earns its keep.
Mem0
Mem0 is an AI memory layer that gives agents and apps persistent context across sessions
Vectorize
Agent memory that learns — Hindsight gives AI agents persistent, reflective memory across sessions
Mnemosyne
Mnemosyne is a local-first AI agent memory layer: one pip install for SQLite-backed persistent recall with sub-millisecond queries.
Featured Head-to-Head Comparisons
Memoryport vs Spider Cloud
Memoryport and Spider Cloud serve very different needs. Memoryport is for users who want persistent, local-first memory across multiple AI coding assistants, ideal for continuity in LLM interactions. Spider Cloud is a web data extraction API designed for AI agents and RAG pipelines needing real-time web content. Your choice depends on whether you need long-term context storage or dynamic web data fetching. Neither is a substitute for the other.
Memoryport vs Temporal Ai
Choose Temporal AI if you need enterprise-grade durable execution for AI agents and long-running workflows, with built-in retries, state recovery, and human-in-the-loop features – it's trusted by OpenAI and Replit for mission-critical orchestration. Choose Memoryport if you're a developer using multiple coding assistants (Claude Code, Cursor) and want a private, local-first memory layer that persists context across sessions without cloud dependency. They solve different problems: Temporal ensures your agents never lose progress; Memoryport ensures they never forget the conversation.
Memoryport vs Presto Voice
Presto Voice and Memoryport serve completely different domains. Presto Voice is a specialized drive-thru AI for QSR chains, boosting revenue through upselling and automation. Memoryport is a developer-focused memory layer for LLMs, ensuring continuity across coding sessions. Choose based on your vertical: restaurant operations vs. AI workflow persistence.
Alternatives to Memoryport
View all
Frequently Asked Questions
Topics
Used Memoryport? Help shape our editorial sentiment research.