We built a tool that installs frameworks like ComfyUI, Ollama, OpenWebUI etc on any cloud GPU in one command and saves your whole setup between sessions [R]

“`html

Summary

A new tool called swm-gpu has been developed to streamline the process of setting up and managing AI frameworks such as ComfyUI, OpenWebUI, Ollama, and others on various cloud GPU providers. The tool automates the installation process and ensures that all custom nodes, models, configurations, and outputs are stored in S3-compatible object storage for easy retrieval between sessions.

Key Features

• swm gpus -g a100 –max-price 2.00 –sort price: Shows the cheapest available GPU across multiple providers like RunPod, Vast ai, Lambda, and others in one view.
• swm pod create: Spins up an instance on any chosen provider.
• swm setup install comfyui: Installs ComfyUI on the created pod.
• The tool includes a workspace sync mechanism. Your entire AI setup, including custom nodes, models, outputs, and configurations, lives in S3-compatible object storage (e.g., B2). When you’re done with your session, running swm pod down pushes everything to the cloud, kills the instance, and ensures that upon restarting on any provider, everything is exactly where it was left.
• Lifecycle guard: Watches GPU utilization. If nothing happens for 30 minutes (configurable), it saves your workspace and terminates automatically, helping users avoid unexpected charges.
• The tool also includes features like a background auto-sync daemon that pushes changes every 60 seconds to ensure no manual saving is needed.
• For large workspaces with many small files, there’s a Tar mode for efficient uploads and downloads.
• Supports additional frameworks such as vLLM, Ollama, Open WebUI, SwarmUI, Axolotl, Cursor, Claude Code, Codex, Windsurf, and more. Users can also use an AI agent like Cursor or Claude Code to manage their GPU instances.
• The tool is free, open-source, and licensed under the Apache 2.0 license. It’s available via pipx installation: pipx install swm-gpu. You can also access it through its website: https://swmgpu.com/ and GitHub repository: https://github.com/swm-gpu/swm.

Key Takeaways

• The tool automates the tedious task of reinstalling AI frameworks on different cloud providers.
• It ensures that all custom setups are saved and easily restored, reducing setup times significantly.
• The inclusion of a lifecycle guard helps prevent unexpected expenses by automatically terminating idle instances.
• The support for multiple AI tools makes it versatile for users working with various models and services.

“`

This HTML document is structured to mirror the original Reddit post, providing a clear overview and key features of the tool while maintaining its structure.

Originally published at reddit.com. Curated by AI Maestro.

Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.

Please enable JavaScript in your browser to complete this form.

Name *

First

Last

Name Email

Email *

Submit