You could use a pseudo r-squared measurement or a measure of accuracy like RMSE, MAE, and so on.

I think taking a look at a plot of predicted values vs. observed values is always a good idea.

Thanks for your help

To check whether a mixed model provides a good fit, consider these criteria:

Fixed effects: Significant p-values indicate significant contributions.

Random effects: Variance should not be close to zero, or random effects may be unnecessary.

Residuals: Should be randomly distributed with no patterns.

AIC/BIC: Lower values indicate a better fit compared to other models.

Likelihood ratio test: Used to compare nested models.

R-squared: Check marginal and conditional R-squared for explanatory power.

Convergence: Ensure that the model has converged properly.

These factors help determine the suitability of the model.