Active since Sep 2025
I tried image analysis with one model, then prompt-based generation with another. The visual and multimodal support fits right into my workflow.