Multiwavelength observations of Galactic black hole transients have opened a new path to understanding the physics of the innermost parts of the accretion flows. While the processes giving rise to their X-ray continuum have been studied extensively, the emission in the optical and infrared (OIR) energy bands was less investigated and remains poorly understood. The standard accretion disc, which may contribute to the flux at these wavelengths, is not capable of explaining a number of observables: the infrared excesses, fast OIR variability and a complicated correlation with the X-rays. It was suggested that these energy bands are dominated by the jet emission; however, this scenario does not work in a number of cases. We propose here an alternative, namely that most of the OIR emission is produced by the extended hot accretion flow. In this scenario, the OIR bands are dominated by the synchrotron radiation from the non-thermal electrons. An additional contribution is expected from the outer irradiated part of the accretion disc heated by the X-rays. We discuss the properties of the model and compare them to the data. We show that the hot-flow scenario is consistent with many of the observed spectral data, at the same time naturally explaining X-ray timing properties, fast OIR variability and its correlation with the X-rays.