Get your model answering requests through MAX.
What it does
Launches an existing model as an OpenAI-compatible HTTP endpoint with MAX and checks that it answers requests coherently.
When to use it
Use when launching or hosting an existing model on MAX, serving a custom architecture, or troubleshooting server startup.
How to use it
Give it a model repository or checkpoint path; it sets up and launches a local model server, then checks its responses.
What you provide
- a Hugging Face repo or checkpoint path
Uses
PI
Latest review
What you need · 1
Serving requires the nightly max binary with max[serve] or max-serve dependencies; the workflow installs an environment if it is missing.
Access · 7
This skill
Hugging Face
Read
Downloads model configuration and weights
https://pixi.prefix.dev/latest/#installation
Read
Follows Pixi installation instructions
pixi
Execute
Installs Pixi when missing