> ## Documentation Index
> Fetch the complete documentation index at: https://docs.clickml.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Model inference component

> Apply a trained ClickML Studio model to your test set or new data to produce predictions and feed results into downstream evaluation steps.

The **Inference** component runs a trained model against the test features (`X Test`) and outputs a DataFrame of predictions. Connect its output to the **Evaluation** component to compute metrics.

## Configuration

No configuration required. Inference automatically uses the model and test data connected to its input handles.

## Input / Output

|        | Type                                             |
| ------ | ------------------------------------------------ |
| Input  | Trained Model + `X Test` (from Train-Test Split) |
| Output | DataFrame of predictions (column: `prediction`)  |

## Notes

* The number of features in `X Test` must exactly match those used during training. If you apply preprocessing after the split, make sure the same steps are applied consistently.
* For multi-output models, the output DataFrame will have columns `prediction_0`, `prediction_1`, etc.