A model card is a short document published with an AI model that explains what it is for, how it was trained and tested, and where it falls short.
In plain words
Model cards were proposed so people could judge a model before using it. A good card covers intended uses, limits and known biases, training data, and evaluation results. On Hugging Face, every model repository's README file is its model card, written in Markdown with extra metadata.
A quick example
Before choosing a speech-recognition model, a developer reads its model card. It lists the languages it was trained on, its error rates on test sets, and a warning that accuracy is lower for some accents. That helps decide whether it suits their users.
Where you'll meet it
- Model pages on Hugging Face.
- New model releases from AI labs.
Related
Sources
- PaperModel Cards for Model Reporting, Mitchell et al. (arXiv) · read 29 Sept 2026
- DocsModel Cards, Hugging Face Hub documentation · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.