mindsdb/minds

Make AI do actual work. Swap the model anytime — keep everything you've built.

What it solves

MindsHub Cowork is a unified workspace designed to move AI from simple chat to actual work. It allows users to delegate entire projects—such as research, analysis, reporting, and scheduled operations—and turn the resulting AI-generated output into shareable web applications, dashboards, and documents without needing engineering skills.

How it works

The platform acts as a "superproject" that integrates several core components:

  • Connected Data: A secure vault that links to external systems like BigQuery, Postgres, Gmail, Drive, HubSpot, Notion, and Linear.
  • Model Router: A system that allows users to switch between proprietary frontier models (e.g., Claude, GPT, Gemini) and open-source models (e.g., DeepSeek, Qwen, Kimi) seamlessly.
  • Open Agents: It uses interchangeable open-source agent harnesses, such as Anton and Hermes, which can be swapped via a dropdown menu.
  • Artifacts: A mechanism to convert agent outputs into live, publishable URLs for apps, code, and dashboards.
  • Infrastructure: It supports flexible deployment across cloud, VPC, on-prem, or air-gapped environments.

Who it’s for

It is built for knowledge workers, including creators, strategists, and operators who need to automate repetitive multi-step workflows and build internal AI tools without writing code.

Highlights

  • Model Agnostic: Swap models anytime while keeping the work you've built.
  • Secure Data Vault: Connects to multiple data sources while keeping credentials scoped and hidden from agents.
  • Publishable Artifacts: Turn AI outputs into live web applications and dashboards.
  • Scheduling & Memory: Includes cross-session memory, a reusable skill library, and the ability to run tasks on a schedule.