ModelBrain fact sheet
ModelBrain is a free, local-only memory layer for AI assistants: it runs its own AI models on your computer to answer questions from your notes, checked against the notes it cites. It works with Claude Desktop, Claude Code, Cursor and ChatGPT, on macOS, Windows and Linux. Every figure below is current as of this page's last update; nothing here is rounded for effect.
At a glance
Release status
Beta on macOS, Windows and Linux. The Linux build is command line only.
Reasoning-based recall-test accuracy
86% by the standard a person reading the answer would call correct, 70% by a strict exact-match standard. 100% on prompt-injection resistance. See "How we tested it" below for the breakdown. Doesn't yet include detecting contradictions between your own notes (see "what it doesn't do yet" on the homepage).
Platforms
macOS (Apple Silicon), Windows (x64 and ARM64), Linux (x86_64 and aarch64, glibc 2.28+)
Price
Free, $0, for one person on their own machines, permanently. No trial, no cap
Protocol
MCP (Model Context Protocol): remember, recall, forget, expand
Works with
Claude Desktop, Claude Code, Cursor (direct local MCP); ChatGPT (via hosted relay)
Company
NexSpark LLC, Atlanta, GA
Platforms
Downloads and SHA-256 checksums: get-it.html.
How we tested it
807 questions, run three times independently and averaged, built to catch the cases that actually break a memory tool: two similar projects you'd confuse, a fact you've since corrected, a long pile of notes with the right one buried in it, a question with no answer in your notes at all, and a deliberate attempt to feed it a bad instruction. "Correct" below means a person reading the answer would call it right; the stricter exact-match number is 70% overall.
Test case
Right answer rate
A single fact you told it once
100%
A prompt-injection attempt in your notes
100%
A question with no answer in memory (correctly declines)
100%
A fact you've since corrected
94%
A long pile of notes, right one buried
90%
Lookalike facts designed to confuse it
90%
Connecting two separate notes to answer one question
79%
Not included above: noticing on its own that two of your notes contradict each other. That's not built yet (see "what it doesn't do yet"), so we don't count it toward this number. Full methodology and the harness itself ship with v1.
Price
Free
The whole product, one person, their own machines, permanently
$0
Device sync
Sync between your own devices, plus managed encrypted backup: coming soon, not built yet
$6/device/mo (billed annually), $8/mo monthly; $4/device past 25
Teams
Shared team vault, SSO, audit trail: designed, not built yet
$6/node/mo, same terms as device sync
Storage
Backup snapshots and any server-held vault; storage on your own computer is never metered
$0.05/GB/mo
Full terms: pricing.html.
Hardware & model tiers
Remembering and plain search run on essentially any machine. Reasoning runs a real local model, so ModelBrain measures free RAM and CPU cores and picks one of these tiers automatically. Nothing to configure, though it can be pinned. How reasoning mode works.
Minimal
7.5 GB free RAM, 4 CPU cores
Granite 3.3 2B Instruct Q4_K_M, ~1.5 GB download
Default
15 GB free RAM, 8 CPU cores
Qwen3 4B Instruct 2507 Q4_K_M, ~2.5 GB download
High
30 GB free RAM, 8 CPU cores
Qwen3 4B Instruct 2507 Q4_K_M, ~2.5 GB download (same model, larger context)
Off
Below 7.5 GB free RAM or 4 cores
No model downloaded; plain search, remember and recall still work
RAM here means free RAM available to ModelBrain, not total installed RAM. Cores means performance cores, not raw thread count. Full detail: hardware requirements & model tiers.
What leaves your computer
Saving, recalling or deleting a memory (Claude Desktop, Claude Code, Cursor)
No
Saving, recalling or deleting a memory (ChatGPT)
In transit only, not stored
Working out what a note means (embeddings)
No, your CPU
Reading a folder of documents you approved
No
Usage tracking, analytics, crash reports
None exist
Managed backup (paid)
Coming soon, not built yet. Will be encrypted first.
Server / team collection (paid, unbuilt)
Yes, readable
Vault encryption at rest
Not today: use your OS's disk encryption; planned as a premium feature
Full detail, including why the ChatGPT row differs from the other three clients: privacy.html.
What it doesn't do yet
Encrypt the vault itself
No: use FileVault, BitLocker, or LUKS
Read scanned or image-only PDFs
No: reads Markdown, plain text, Word, PowerPoint, Excel, and text-layer PDFs
Detect contradictory notes on its own
No: you tell it which note is current
Team sharing
Designed, not built: won't ship before per-person permissions exist
Device sync and managed backup
Coming soon: priced on this page, but not built yet