Boogu-Image-0.1-Turbo is a four-step distilled text-to-image diffusion model developed by Boogu that generates high-quality 1024×1024 images from text prompts, containing 10 billion parameters trained through joint training with 25-50 inference steps. Built on a unified multimodal architecture, it prioritizes photorealistic output and bilingual (Chinese-English) text rendering under the Apache 2.0 license. The model achieves strong performance using training data roughly one order of magnitude smaller than competing closed-source systems. Best use cases include realistic photography with embedded text, poster and product design, stylized and creative generation, and rapid iteration in production with four-step inference. The Turbo variant delivers four-star performance on realistic photography, solid four-star simple text rendering, and three-star dense text rendering. Limitations include noticeably weaker performance on dense text rendering compared to its Base variant, showing three-star performance versus four stars on the Base model.