Google Launches Android Bench 2.0 for Real Android Development

Google introduced Android Bench 2.0, a benchmark evaluating model+agent combinations on real Android engineering tasks from Google best practices. Top performers were GPT-6 Astra with Codex and Claude Fable 5.1 with Claude Code, but the highest pass rate was only 28%. Many tasks, including Flutter-to-Android and XML-to-Compose migrations, saw zero passes across all models.