Hub.ai

Hub.ai is a platform designed to rapidly deploy applications directly from public GitHub repositories or from a description of your desired software environment. It provisions an isolated Virtual Machine (VM), installs necessary dependencies, configures the software, and provides you with a running instance accessible via a unique URL, often within seconds.

Instead of providing a GitHub URL, you can simply describe the software stack you need (e.g., "NodeJS 18, Redis, PostgreSQL 15"). Hub.ai's AI interprets your request, selects an appropriate base image, installs and configures the specified software on a new VM, and provides you access to the running environment.

Hub.ai is ideal for developers, testers, educators, and anyone who needs to quickly spin up environments for:

• Demoing projects or features.
• Testing pull requests in isolation.
• Exploring open-source projects without complex local setup.
• Creating temporary environments for workshops or presentations.
• Quickly deploying and testing backend APIs or AI/ML models.

Deployment failures can happen for various reasons, such as missing dependencies, incorrect build commands, or application errors during startup. You can view the deployment logs on the specific deployment's detail page to help diagnose the issue. Common problems include:

• Repository missing necessary configuration files (e.g., `package.json`, `requirements.txt`).
• Application trying to bind to a port already in use or requiring specific environment variables not set.
• Build process failing due to code errors or incompatible dependencies.

If you continue to have trouble, check the repository's documentation or contact our support team via the Contact Us page.

You can sign up or sign in using your existing GitHub or Google account. Visit the Sign In / Sign Up page and choose your preferred provider.

Hub.ai uses a simple pay-as-you-go credit system. 1 credit = $1.00 USD.

You are charged based on the resources (CPU, RAM, Disk, optional GPU) consumed by your VM instance, billed per minute only while the instance is actively running. Stopping an instance immediately stops credit consumption for that instance.

Your very first instance launch costs a one-time fee of $0.99, which also grants you $10.00 worth of starting credits.

You can see estimated hourly costs for different instance sizes on the homepage and manage your credits in the Billing section of your account.

You can purchase additional credits at any time through the Billing section in your account settings. We offer several predefined credit packages, or you can specify a custom amount.

No, purchased credits do not expire.

Your current credit balance and detailed billing history, including charges per instance, are available in the Billing section of your account.

Currently, Hub.ai primarily supports public GitHub repositories. The platform attempts to automatically detect the project type and necessary build/run steps. Support for private repositories may be added in the future.

Hub.ai uses heuristics and AI analysis to detect common project structures and configuration files (like `package.json`, `requirements.txt`, `docker-compose.yml`, `Dockerfile`, etc.). It attempts to install dependencies and start the application based on standard conventions for the detected framework or language. For complex or non-standard setups, you might need to ensure your repository includes clear startup instructions or a suitable configuration file.

We offer a range of VM sizes, from Micro (1 vCPU, 1GB RAM) to XX-Large (16 vCPUs, 64GB RAM) and beyond. High-performance GPUs (like RTX 4090) are available for larger instances (typically requiring 24GB+ RAM). You select the desired resources when creating a VM or launching a deployment. See the homepage for sample configurations and costs. Custom configurations are also available via contacting sales.

Direct SSH access is generally not provided as instances are managed. However, we offer an interactive web terminal for many instances, accessible through the VM details page, allowing you to run commands within the running environment.

You can run multiple instances simultaneously, limited primarily by your available credits. You can deploy the same repository multiple times or deploy different repositories concurrently.

Data stored on the VM's local disk is generally persistent as long as the VM exists. However, VMs are designed for rapid deployment and testing; they are not intended for long-term persistent data storage or production databases without proper backup strategies. Stopping and starting a VM retains its disk state. Deleting a VM permanently removes its disk and all associated data.

You can view, start, stop, and delete your VM instances from the VMs page. Deployment configurations and logs can be viewed on the Deployments page.

Yes, Hub.ai has support for repositories containing a `docker-compose.yml` file. If detected, it will attempt to build and run the services defined in your Compose file. Ensure your `docker-compose.yml` is configured correctly for the environment.

Currently, Hub.ai primarily focuses on exposing a single main web application port (like 80, 443, 3000, 8000, 8080). Support for exposing multiple custom ports is limited. If your application requires multiple ports to be accessible externally, it might not function as expected in the default deployment environment.

To update an application deployed from a GitHub repository, you typically need to:

1. Push your code changes to the GitHub repository.
2. Create a new deployment from the updated repository/branch via the Hub.ai interface.

Hub.ai does not automatically track repository changes and update running instances. Each deployment creates a new, independent instance based on the repository state at the time of deployment.