Cloning with instant voice cloning can be complicated, and we have some general guidelines. However, they are just that: guidelines. We don't have any set rules regarding the number of samples or length. We've seen users use samples of only 30 seconds and get excellent results, while we've also seen some users use 10 minutes of audio and have worse results. But we do have a few things that you should consider.
Audio quality is the most important aspect to consider when using instant voice cloning.
The number of samples is irrelevant; what's important is the total run time. Having more than 2-3 minutes of audio will yield little improvement and can, in some cases, even be detrimental to the stability of the clone.