Flagship · early access
MinniMemoryMCP
An MCP server that shrinks what Claude Code loads every turn
Checks your agent’s memory files and splits them into a small always-on part and files that load only when needed. Every line is kept.
NewMinniMemoryMCP is in early access
Reduce your coding agent’s token usage with MinniMemoryMCP.
Now available in Claude Code
MinniMemoryMCP splits your agent’s memory into a small core that loads every turn and files that load only when a task needs them. Measured in Claude Code: 27% cheaper sessions, no lines lost.
$ npx -y minnimemory install
Before: the memory that loads every turn, in full
stable rules and conventionsdated notes and task detail
Demo
Watch Minni AI at work.
Our research and four tools in one video: MinniMemoryMCP, MinniCode, MinniTuner, and MinniMemoryAgent, a memory upkeep agent that is not public yet. Every number on screen comes from a real run of our tools.
Evidence
How we measured 27.4% cheaper Claude Code sessions.
Four results, measured. Pick one to see what it means.
Our final result: 27.4% cheaper
Across 81 answers, the bill came out 27.4% smaller once MinniMemoryMCP sorted the memory. A blind grader marked 72 correct after, 74 before.
Before3,574,009
After3,019,536
tokens, 81 answers per side · Claude Sonnet 5.5 · Claude Code 2.1.292| Tokens | Before | After | Change |
|---|---|---|---|
| Cache writes | 626,130 | 391,988 | -37.4% |
| Cache reads | 2,931,482 | 2,609,556 | -11.0% |
| Output | 16,181 | 17,752 | +9.7% |
| Uncached input | 216 | 240 | +24 |
| Total tokens | 3,574,009 | 3,019,536 | -15.5% |
| Cost | $2.31 | $1.68 | -27.4% |
Source: MinniMemoryMCP 1.1.0, Claude Sonnet 5.5 in Claude Code 2.1.292, 81 answers per side, measured 6 October 2026
How we measured
Over 3.5 million tokens of measured work
After many rounds of testing and tweaking our interface, this is our final run. We gave Claude Sonnet 5.5 the same work in Claude Code, three sessions a side and 81 answers in all: once with the test memory as it was and once after MinniMemoryMCP sorted it. The run used 3,574,009 tokens before and 3,019,536 after, 15.5% fewer, and cost $2.31 before and $1.68 after, 27.4% less. A blind grader marked 74 of 81 answers correct before and 72 of 81 after.
MinniMemoryMCP 1.1.0, measured 6 October 2026
Less memory loaded on every turn
Our full-size test memory went from 12,511 tokens to 6,699 tokens loaded on every turn. The rest moves into files the agent opens only when a task needs them. No lines are lost.
Counted with MinniMemoryMCP 1.2.1 on a full-size test memory, 10 October 2026
Every line kept, every file restored
We compiled the same 16-file test memory and the example file that ships with the package. Every line landed somewhere word for word, and undo put every file back byte for byte.
MinniMemoryMCP 1.2.1, measured 9 October 2026
Shorter answers with our answer-style rules
We asked an AI agent the same questions with and without the rules, 240 answers in all. With the rules, answers used 58.1% fewer tokens.
Measured 5 September 2026
Methods
Session cost: the tokens of every call, priced at Anthropic's published API rates, with the cache reads and writes each call made. Cost falls further than tokens because the saving is mostly in cache writes, the most expensive input, while cache reads, most of the tokens, cost a tenth as much. The grader saw each pair of answers blind, in both orders, and a verdict counted only when both orders agreed.
Memory counts: the token counter built into MinniMemoryMCP 1.2.1, run on a full-size test memory of 16 files loaded every turn. Anyone with the package and the same files gets the same numbers.
The 58.1%: the same questions answered with and without the rules, counting the tokens in each answer.
The service
The MinniMemory Interface: rules for a smaller agent memory.
Our research, written down: how an agent’s memory is split, what it reads every turn, and the rules that keep its answers short. Every result above comes from it. You can apply it two ways today.
Do it yourself
MinniMemoryMCP
Free on npm. Run it in your project and it applies the interface to your agent’s memory files. Every line is kept, and undo puts them back.
Open MinniMemoryMCPDone with you
We apply it to your agent
We measure what your agent is handed each turn, apply the interface to its memory and instructions, and report tokens saved against answer quality.
More ways to apply it are on the way.
Products
Three tools, each free to try.
Beta
MinniCode
A free AI coding agent that runs on OpenRouter models
Reads, edits, tests and opens pull requests in your project, on free AI models. Bring a free OpenRouter key, one line to install.
See how it works ↓Open page
Beta · Windows 11
MinniTuner
A Windows 11 optimizer that shows every change and can undo it
See what your PC is doing, change only what you tick, and undo it. BasicTuner is free.
See how it works ↓Open page
How it works
How each product works.
Maps the memory files your agent loads and flags what does not need to load on every turn.
Open the MinniMemoryMCP page$ minnimemory init 16 files load every turn CLAUDE.md, MEMORY.md + 14 linked files Open items, current state change often, not needed each turn Session logs history, only needed when asked Plan: 5,812 tokens load on demand Per turn: 12,511 -> 6,699 tokens (-47%)
Support the research
Help keep the research running.
MinniMemoryMCP and MinniCode are free, and so is BasicTuner. We pay for the test runs, the AI calls and the hosting behind them. If our tools saved you time or tokens, you can back the work.
Payments go through Buy Me a Coffee. Support is a gift: it does not buy a product or a license, and it is not tax deductible.
Before you install
Frequently Asked Questions
01Do you see my code or my memory files?
MinniMemoryMCP never does. It runs entirely on your computer, sends nothing anywhere, needs no account or API key, and has no AI model inside it. MinniCode sends each request through our servers to the OpenRouter model you picked, on your own key, and we don't store those requests. The chat window on this page is answered by an outside AI provider, and we don't store those messages either.
02Why not just edit the file myself?
You can, and for a short file you should. It gets harder as the file grows: you have to decide which lines your agent needs on every turn, move the rest into separate files without dropping anything, and leave a pointer so the agent knows when to open each one. MinniMemoryMCP does that sorting for you, backs up your originals first, and can re-sort after you edit the file again.
03What if it wrecks my memory files?
It backs up your originals before it writes anything, and one command, undo, puts them back exactly as they were. Inside Claude Code the server itself never writes: apply and undo return the command to run and the list of files it touches, and your agent runs that command through your own approval prompt, so you see every change. Run from the terminal, apply writes the files itself, after the backup.
04What does it cost?
MinniMemoryMCP is free while it is in early access, and launch pricing is not set yet. MinniCode is free. MinniTuner is in open beta: BasicTuner is free, and Advanced scans for free but needs a license key to apply its changes.
05Who makes Minni?
Three software engineers from big tech. We ran into these problems in our own workflows: coding agents re-reading their whole memory every turn, and token bills growing with it. Minni is what we built to fix that for ourselves first.
Talk to us before it gets expensive.
Send us what your agent loads each turn. We apply the interface and report tokens saved against answer quality.