A confidence interval of the prediction provides a range of values for the mean response associated with specific predictor settings. For example, for a 95% confidence interval of the prediction of [7 8], you can be 95% confident that the mean response will fall within this range.
The prediction interval is always wider than the confidence interval because of the added uncertainty involved in predicting a single response versus the mean response.