Al Rassam Al Arabi V3.1 R1 37 -
As of the latest release, access is available through three primary channels:
System Requirements (for local deployment): Al Rassam Al Arabi V3.1 R1 37
Working with Multilingual Documents:
Design and Layout:
For years, AI image generators have produced garbled, nonsensical "gibberish" when asked to write Arabic words. This model solves that. It can embed accurate, readable Arabic phrases into images—whether it’s a bismillah on a manuscript or a shop sign in a generated street scene. As of the latest release, access is available
The model includes a built-in "Urf" (custom) filter. Unlike Western models that block arbitrary concepts, this filter understands cultural nuances. For instance, it will not generate pork products or idolatrous imagery but will accurately generate a majlis (sitting area) or a qibla wall with correct orientation. System Requirements (for local deployment):
Al Rassam Al Arabi V3.1 R1 37 supports a native resolution of 1024x1024 with upscaling capabilities to 4K. Its context window allows for multi-turn generation, meaning you can refine an image using conversational Arabic commands without starting from scratch.
