Hugging Face has introduced QIMMA (قِمّة), a new leaderboard designed specifically to evaluate Arabic language models with a quality-first approach. The initiative addresses a gap in the AI evaluation landscape, where Arabic language models have historically received less attention and rigorous benchmarking compared to their English counterparts. QIMMA establishes standardized quality metrics and evaluation criteria tailored to the linguistic and cultural nuances of Arabic, enabling developers and researchers to better assess model performance across Arabic-speaking regions. The leaderboard represents a significant step toward more equitable AI development and deployment across non-English languages. By creating transparent, community-driven benchmarking standards for Arabic LLMs, QIMMA aims to encourage higher-quality model development while providing clearer guidance for organizations looking to implement Arabic language AI solutions. The initiative reflects broader efforts within the AI community to ensure that language model improvements extend beyond English-dominant markets.