modeldock

Introduction

ModelDock launches Hugging Face models behind an OpenAI-compatible API. Paste a model link, and Pilot, an AI agent, checks the model, picks a GPU and deploys it. You then call it with the OpenAI SDKs by changing the base URL and the model name; the rest of your code stays the same.

base URL
https://api.lab.litepod.io/v1

How it works

  • Launch. Sign in with your litepod account and paste a Hugging Face model link in the dashboard. Models up to 10B parameters are supported.
  • Pilot deploys it. Pilot checks the model, picks the smallest GPU that fits and starts it. The model has to cold-start before it can answer.
  • Call it. Create an API key and use the model id ModelDock gives you with any OpenAI SDK.

Free test launches (Pre-flights)

Every launch is a free Pre-flight today: the model stays up for 15 minutes or 25 requests, whichever comes first, then it is torn down. See Errors and limits for the exact caps and Billing and pricing for how paid launches are priced.

Where to go next

  • Quickstart: your first model and first call.
  • OpenAI compatibility: supported endpoints and fields.
  • Models: popular Hugging Face models you can launch, with size, license and context length.
  • Pricing: what each GPU tier costs per million tokens.

More from litepod: the free LLM API with free open-source models. ModelDock is a litepod product.