Newly Features
We've cooked up a bunch of improvements designed to reduce friction and make the.
Create ->
Runpod is happy to announce the release of the Hub — a comprehensive platform designed to get you started with serverless applications focused on AI. Whether you're deploying a fine-tuned model or spinning up a custom workflow, the Hub gives you a head start. No manual setup. No Docker registry uploads. Just deploy, run, and go.
Ready-to-Deploy AI Solutions
The Hub makes it easy to deploy powerful, community-vetted repos — already configured to run on Runpod’s high-performance serverless infrastructure.
ComfyUI - A node-based interface for Stable Diffusion that lets you build, remix, and run image generation workflows with full visual control.
Whisper - A fast, high-accuracy speech recognition system powered by OpenAI’s Whisper. Great for audio pipelines and real-time transcription
Mochi 1 - A groundbreaking open-source AI model that's changing the game in video generation with smooth, realistic motion at 30 frames per second.
And that’s just the start. You’ll find tools for LLM inference, fine-tuning, embeddings, agents, and more — all runnable in just a few clicks.
Community-Driven Innovation
The Hub isn’t just built by us — it’s built by you.
We’ve opened the door for community contributions, so anyone can publish a public repo and make it instantly deployable
We're already seeing exciting contributions from the community:
Axolotl - A powerful tool for fine-tuning large language models, contributed by the Axolotl AI Cloud team.
Ollama - We've worked directly with individual developers to bring Ollama's local LLM capabilities to the Hub.
This collaborative approach brings you closer to the people that make AI happen, ensuring the Hub continues to grow with the latest innovations in AI technology.
This is how open-source AI should feel: remixable, runnable, and built in the open.
Seamless GitHub Integration for Builders
Tired of building containers, uploading to Docker registries, and writing deployment scripts? Us too.
When you contribute your work to the Hub, you can deploy directly from a GitHub repo. Just write your Dockerfile, push your code, and we’ll take care of the containerization and provisioning behind the scenes.
No config headaches. No infra blockers. Just deployment that works.
Getting Started is Simple
Spinning up a Hub listing is as simple as:
- Browse and select your desired listing from the Hub
- Click Deploy
That’s it. Within moments, you’ll have a live endpoint running on your account — fully serverless, automatically scalable, and ready to integrate with whatever you’re building.
Open Source Collaboration
All Hub listings are open source and GitHub-connected, which means:
- You can fork and customize any repo
- You can submit issues or pull requests directly to the original project
- You don’t need to guess how anything works
The feedback loop between builders and users is built right in — so things improve fast, and you can ship even faster.
Ready to Contribute?
We’d love to see it. Check out the Hub documentation for details on how to contribute your own repo. It’s fast, free, and a great way to share your work with other AI developers.
Whether you're building something new or just need a faster way to run open-source AI tools, the Hub gets you there faster — without making you do the dirty work.
Explore the Runpod Hub → Launch the Hub
.webp)
.svg)