Avian influenza virus (AIV) currently causes a panzootic with extensive mortality in wild birds, poultry, and wild mammals, thus posing a major threat to global health and underscoring the need for efficient monitoring of its distribution and evolution. We here utilized a well-defined AIV strain to systematically investigate AIV genetic characterization through rapid, portable nanopore sequencing by comparing the latest DNA and RNA nanopore sequencing approaches and various computational pipelines for viral consensus sequence generation and phylogenetic analysis. We show that the latest direct RNA nanopore sequencing updates improve consensus sequence generation, but that the application of the latest DNA nanopore chemistry after reverse transcription and amplification outperforms such native viral RNA sequencing by achieving higher sequencing accuracy and throughput. We additionally leveraged the direct RNA nanopore sequencing data for the detection of RNA modifications, such as N6-methyladenosine and pseudouridine, which play a role in viral immune evasion. Finally, we applied these sequencing approaches together with portable AIV diagnosis and quantification tools to environmental samples from a poultry farm, demonstrating the feasibility of nanopore sequencing for on-site non-invasive AIV monitoring in real-world outbreak scenarios.