Xai has introduced an update to Grok’s image generation capabilities with the release of a new autoregressive model, codenamed Aurora. Available on the š¯•¸ platform, the Aurora model is designed to provide photorealistic rendering, image synthesis for precise text, and advanced multimodal access capabilities. The update is initially available in select countries and will roll out to all users within a week.
Aurora: A Breakthrough in Imaging Generation
Aurora is an avororegressive hybrid network that processes billions of samples, text and image data from the web. This extensive training provides Aurora with the ability to create both visual and textual elements, high-quality images, follow complex instructions, and perform user-provided visual editing.
The model offers advanced features such as:
- Photorealistic renderingcapable of producing images with detailed real-world accuracy.
- Multimodal access supportAllows for seamless image editing and inspiration based on user-supplied visuals.
- Accuracy in artistic domainsincluding logos, memes and real human portraits.
Examples of Aurora’s capabilities include realistic landscapes such as artistic visuals, creative logos, and a Cybertruck under Aurora or abstract artistic representations of complex instructions.
High quality image generation in domains
Aurora addresses issues that other image generation models often face:
- Enterprise woundproduces accurate images of real-world objects and celebrities.
- Artistic text and logo creationallowing for detailed and visually appealing designs.
- Real portraitspresenting lifelike representations of people in different styles.
Advanced image editing features
In addition to creating images from scratch, Aurora provides powerful image editing tools. Users can now upload images for modification, offering more flexibility and control over their creative projects. This feature will soon be available to all users on the š¯•¸ platform.
For example, users can turn a photo of a cat into an anime-style image or apply advanced stylistic adjustments to existing visuals.
Photo: Khai