Google updates Android Bench with new LLMs, but Gemini still lags behind

Google updated Android Bench in July 2026 by adopting the Harbor framework and adding eight new models including Claude Fable 5, Claude Sonnet 5, and others to its LLM leaderboard for Android development tasks. The refresh moved older Gemini models to archived status, including Gemini 3 Pro Preview and earlier versions, while introducing Gemini 3.5 Flash. The title's reference to Gemini lagging suggests that while Gemini models historically led the benchmark, newer competitive models may have closed the performance gap with this update.

Source: Ars Technica | 2026-07-08