This is a pre-trained bilingual large language model with 130 billion parameters, supporting Arabic and English. It is trained on a dataset of 72 billion Arabic tokens and 279 billion English/Code tokens. The Arabic data has been iterated 1.6 epochs compared to 1 epoch for English/Code, with a total of 395 billion tokens trained. The model is based on the Transformer decoder-specific architecture (GPT-3), utilizing the SwiGLU nonlinear activation function. It implements ALiBi positional embeddings, allowing extrapolation to long sequence lengths, providing enhanced context processing and model accuracy.