Pro, an updated version of its multimodal model, Janus. The new model improves training strategies, data scaling, and model size, enhancing multimodal understanding and text-to-image generation.

DeepSeek has upended the AI industry over the last week, but its Janus Pro image model won't have Midjourney or others worried just yet.

The Chinese startup has released an image-generation model called Janus-Pro, aiming to take on US rivals DALL-E 3 and Stable Diffusion. The multimodal model, which generates images from text prompts, is said to outperform competing services in areas such as image quality and accuracy.