GBrain

About GBrain

GBrain is a shared memory and account layer for AI tools, launched by a team that includes Y Combinator's CEO. It stores notes as a folder of markdown files that multiple AI assistants-ChatGPT, Claude, Claude Code, Cursor-can read and write to. The tool also centralizes connected accounts like Gmail and Google Calendar so any connected AI can access them without exposing API keys or passwords directly.

Review

GBrain addresses a concrete problem: switching between AI tools means losing the context and project history each one accumulated. Instead of each AI keeping its own siloed memory, GBrain externalizes that memory into portable markdown files. The tool also handles account connections once, then exposes them to multiple AI harnesses through a single endpoint or CLI.

Key Features

• Shared markdown memory: A folder of plain text files that ChatGPT, Claude, Claude Code, and Cursor all read and write. You can open, correct, or export the files at any time.
• Unified account connections: Connect Gmail, Google Calendar, and other accounts once. GBrain encrypts and stores the credentials, then makes them available to any connected AI harness without revealing keys.
• Team workspaces: Multiple people work in the same thread with one assistant, so no one needs to re-explain a project someone else already briefed the AI on.
• Activity log and rate limiter: A log shows which accounts and tools each AI harness calls. A built-in rate limiter can cut off access if an agent goes haywire.
• Model selection and scheduled jobs: You pick which model runs and can swap later without losing memory. Pre-built jobs-like a daily brief on upcoming meetings-run on a schedule and can pull from Github, Gmail, Stripe, and Mercury.

Pricing and Value

GBrain costs $199 per month for a team workspace. Through September 30, a Product Hunt promotion reduces the first month to $99 and includes $100 of AI usage. The tool doesn't integrate with existing ChatGPT or Claude subscriptions yet; you either bring your own API key or use the included usage credits. Self-hosting the core memory component is possible since the source code is open, but the rest of the stack is not self-hostable at this time.

Pros

• Memory lives in portable markdown files you control, not locked inside any single AI platform.
• Accounts get connected once and work across multiple AI tools, with an activity log for visibility.
• Team workspaces let groups share context without repeating project explanations.
• Connection expiration dates let you grant limited-time access when trying a new AI harness.
• Source code is available on GitHub for the core memory component.

Cons

• Does not work with existing ChatGPT or Claude subscriptions-you need separate API keys or included credits.
• Only the memory component is open source; the rest of the stack requires GBrain's hosted infrastructure.
• Not well suited for individual users who stick with one AI tool and don't need cross-platform memory or team context sharing.

GBrain fits teams and developers who regularly switch between AI coding assistants and chatbots, or who want a single place to manage what multiple AIs know about their projects. The markdown-based memory and centralized account layer reduce the friction of jumping between tools, though the separate API key requirement means it doesn't yet integrate with subscriptions users may already pay for. For teams already coordinating across several AI tools and shared accounts, the workspace model and activity logging give a practical way to track what AI agents are doing.

Open 'GBrain' Website