Open Source Software (OSS) projects often label their open issues, which supports contributors in choosing a task. However, labeling issues is a time-consuming task and many projects face inconsistencies with manual labeling. Many automatic approaches to creating labels are limited to classifying issues as bug/non-bug. As APIs have been explored to provide recommendations, offer tips, or define skills for developers and projects, we investigate the feasibility of automatically labeling issues based on the APIs used in the files that will be changed to complete the tasks. We found that API labels can be predicted with precision up to 77.8%, and recall up to 95.9% that these labels are mainly relevant for experienced newcomers. We also ran an experiment to assess the relevancy of these labels for contributors choosing an issue to contribute to. We asked 74 participants to select three issues that they would choose and report which regions of the issues page supported their decision, and why. We also asked which labels are relevant. The results showed that the issue title and the labels are the most important regions. Finally, The API labels we generated were more selected than the component labels used by the practitioners and expert coders. Our results may help to create tools to automatically tag issues improving the context information over the type of the issue. This will facilitate those developers willing to find tasks that better match their skills.