Getting started

Mystic helps companies deploy and manage Artificial Intelligence (AI) models in the cloud. You can run models either on our serverless endpoints running on our GPUs or you can authenticate Mystic with your cloud account to run there instead.

👍

Pick a section that's relevant to you!

Docs can be pretty overwhelming so pick a section that best describes why you're here to find materials that we have found are most useful for what you're looking for

I'm just trying to find out more about Mystic

Great! Checkout some of the features and guides our users love:

• Fractional GPUs & GPU sharing
• Deploy Mistral 7B with vLLM
• Inference
• FAQ
• Transcribe YouTube videos
• Pricing

I want to run a model on the explore page

Using an already existing model is really simple, check out the following to get your app integrated today on our serverless endpoints:

• Authentication
• Runs
• Files

You might be able to run the model even faster and at a more affordable rate by using our cloud integration, check out the following:

• Cloud integration
• Scaling configuration
• Fractional GPUs & GPU sharing

I want to run a custom model

We have a very basic python SDK that allows you to wrap existing codebases to be run on Mystic, check out the following:

• Pipeline building quickstart
• In-depth pipeline building
• Pointers
• Inputs and outputs
• Fractional GPUs & GPU sharing

I have a problem, how can I fix it?

Bugs are annoying and we're here to help. If somethings broken we always recommend checking out some of our troubleshooting docs as the first step:

• Troubleshooting
• FAQ

If nothing in the above helps we have an active discord community where you can open a private support ticket to speak to our team directly. Our team operates in the UK and is normally most responsive 8am-10pm UTC, but may well respond outside of those hours. Join our discord here.