humain-m3: Saudi Arabia's Arabic-First Frontier Model, Explained
Saudi HUMAIN's humain-m3 scores 89.37% on Arabic benchmarks, edging GPT-5.6 Sol in the company's own eval. Trained on a trillion Arabic tokens; open weights promised October 2026.
Saudi HUMAIN's humain-m3 scores 89.37% on Arabic benchmarks, edging GPT-5.6 Sol in the company's own eval. Trained on a trillion Arabic tokens; open weights promised October 2026.
Start typing to search