This dataset contains the gene annotations generated using BRAKER3 for high-quality de novo genome assemblies of Apis andreniformis (black dwarf honey bee) and Apis florea (red dwarf honey bee). The annotation was performed using a combination of RNA-seq data (long- and short-read) and protein homology evidence. Files included: Apis andreniformis A_andreniformis_braker.gtf – Gene annotation file with exon-intron structures. A_andreniformis_braker.aa – Predicted protein sequences (n=16,479). A_andreniformis_braker.codingseq – Predicted coding sequences (n=16,479). A_andreniformis_braker_longest_isoform_only.aa – Predicted protein sequences with only the longest isoform per gene (n=12,189). Apis florea A_florea_braker.gtf – Gene annotation file with exon-intron structures. A_florea_braker.aa – Predicted protein sequences (n=16,511). A_florea_braker.codingseq – Predicted coding sequences (n=16,511). A_florea_braker_longest_isoform_only.aa – Predicted protein sequences with only the longest isoform per gene (n= 12,207). These annotations are provided as a resource for comparative and functional genomics studies of the Micrapis honey bee subgenus. The associated genome assemblies are available on NCBI GenBank under accession numbers JBLZQI000000000.1 (Apis andreniformis) and JBLZQJ000000000.1 (Apis florea). All raw sequencing data used for genome and transcriptome sequencing are available on the NCBI Sequence Read Archive (www.ncbi.nlm.nih.gov/sra) under the following BioProjects:Apis andreniformis1. Genome sequencing and assembly; NCBI BioProject PRJNA12170362. Transcriptome sequencing; NCBI BioProject PRJNA1217108Apis florea1. Genome sequencing and assembly; NCBI BioProject PRJNA12170412. Transcriptome sequencing; NCBI BioProject PRJNA1217110