Manifold: An open-source tool to run AI models for 90% less

Manifold is a free, open-source toolkit that makes deploying, managing, and securing any AI model as simple and reliable as using Docker.

Stop wrestling with complex cloud configurations and unpredictable costs. Manifold provides the essential infrastructure layer for the AI economy, giving you control over your models.

Deploying AI is chaotic, expensive, and insecure. Developers spend more time on infrastructure than on innovation. Manifold solves three key problems:

• 🤯 Complexity: Months of learning Kubernetes, cloud provider tools, and networking.
• 💸 Cost: Unpredictable, runaway bills from GPU instances and API calls.
• 🔒 Trust: Lack of security, auditing, and governance for model inputs and outputs.

Manifold is the Operating System for AI Deployment, built on three pillars:

1. Simplicity: A single, intuitive CLI to run, list, and stop any AI model as a containerized service.
2. Trust: Every action is captured in a structured, JSON-based audit log, providing a complete, immutable system of record.
3. Control: Built-in (simulated) cost estimation for every transaction, giving you immediate insight into your spending.

• Docker Desktop installed and running.
• Python 3.8+

1. Clone the repository:

   git clone https://github.com/Esrbwt1/manifold.git cd manifold

2. Set up a virtual environment and install dependencies:

   python3 -m venv venv source venv/bin/activate # On Windows: .\venv\Scripts\activate pip install -r requirements.txt

The Manifold system consists of two parts: a background Daemon and a CLI Client.

1. Start the Daemon: In one terminal, start the Manifold daemon. It will run continuously to manage your services.

   uvicorn manifold.daemon:app

2. Use the Client: In a second terminal (with the virtual environment activated), you can now use the manifold CLI.

   • Run a service:

     # Run a simple web server for testing python manifold/cli.py run nginxdemos/hello --name test-server --port 8080:80

     ✅ Daemon acknowledged. Service 'test-server' is starting.

   • List running services:

     python manifold/cli.py ps

     ┌─────────────┬─────────┐ │ NAME │ STATUS │ ├─────────────┼─────────┤ │ test-server │ running │ └─────────────┴─────────┘

   • View the audit log:

     python manifold/cli.py logs test-server

     Audit trail for test-server... 2025-06-30T00:00:00 | $0.000123 | 172.17.0.1 - "GET / HTTP/1.1" 200...

   • Stop a service:

     python manifold/cli.py stop test-server

     ✅ Service 'test-server' stopped successfully.

Phase 1: MVP Complete

• Core service lifecycle management (run, ps, stop).
• Client-Daemon architecture.
• Structured JSON auditing for all container output.
• Simulated cost estimation framework.

This project is under active development. Join us in building the foundational infrastructure for the AI revolution!