Vidu

Vidu is a reference-to-image model that creates brand-new images by combining your reference photos with a written description. Instead of generating from text alone, Vidu looks at the visual subjects you provide and recreates them faithfully inside a fresh scene shaped by your prompt. This makes it an ideal tool for creative professionals who need to keep a character, product, or object looking consistent across multiple images — something that traditional text-only image generators often struggle to achieve.

At its core, Vidu solves a common creative problem: maintaining a recognizable subject from one image to the next. You supply one or more reference images that show the appearance you want to preserve, then write a prompt describing the new situation, action, or setting you'd like to see. Vidu blends the two, producing an image where your subject keeps its distinctive look while being placed into the scenario you've imagined. For example, you might provide reference images of a small character and prompt it to be 'looking at the apple on the beach and walking around it,' resulting in a new image that keeps your character on-model while introducing the beach environment and the described action.

The ability to use multiple reference images at once is one of Vidu's standout strengths. By feeding it several views or examples of the same subject, you give the model more visual information to work with, helping it understand the subject from different angles or in different states. This leads to more reliable and consistent results, which is especially valuable when you're building a series of images that all need to feel like they belong together.

Vidu is a natural fit for a wide range of creative roles. Illustrators and character designers can keep a mascot or original character recognizable across an entire collection of artwork. Marketers and brand teams can place a product into many different scenes and contexts without losing its identity. Storytellers, comic creators, and concept artists can develop a cast of consistent characters and drop them into new environments scene by scene. Content creators who need cohesive visual sets — whether for social media, presentations, storyboards, or campaign assets — benefit from the model's focus on subject consistency.

When it comes to framing your work, Vidu gives you control over the shape of your output. You can choose a widescreen 16:9 format that works beautifully for landscape compositions, cinematic scenes, and horizontal banners; a vertical 9:16 format ideal for mobile-first content, stories, and portrait compositions; or a square 1:1 format that's perfect for social posts, avatars, and balanced centered layouts. The widescreen 16:9 option is the default, making it easy to get started with a versatile, commonly used frame.

Your prompt is where you describe the scene, action, mood, and setting you want. Prompts can be quite detailed — up to a generous length — giving you room to specify exactly what your subject is doing, where it is, and how the scene should feel. The more clearly you describe the situation, the better Vidu can place your referenced subject into a believable and compelling new image. Because the model leans on your reference images for the subject's appearance, your prompt can focus on everything else: the environment, the activity, the atmosphere, and the storytelling details that bring the picture to life.

Vidu also offers a way to influence the randomness of generation through a seed setting. By keeping the same seed, you can reproduce or revisit a particular result, which is helpful when you want to make small adjustments while preserving the overall look you've achieved. Changing the seed lets you explore fresh variations of the same idea, giving you a simple way to iterate and experiment until you land on the version you love.

The model outputs a single finished image per request, delivered as a high-quality image file you can download and use in your projects. Because Vidu is designed around preserving subject identity, the resulting images tend to keep your reference subject looking faithful to the source while seamlessly integrating it into the new scene described by your prompt.

To get the best results from Vidu, it helps to provide clear, high-quality reference images that show your subject well. Supplying multiple references can improve consistency, especially if they capture the subject from different perspectives. Pairing those references with a descriptive, specific prompt gives the model the strongest foundation to work from. Choosing the right aspect ratio for your intended use — vertical for mobile, widescreen for cinematic scenes, square for social — ensures your image fits its destination from the start.

It's worth keeping in mind what Vidu is built to do. This is a reference-to-image tool: it requires at least one reference image and a text prompt to function, so it's not intended for generating images purely from text with no visual starting point. Its specialty is taking subjects you already have and reimagining them in new contexts while keeping them recognizable. For anyone whose work depends on visual continuity — recurring characters, branded products, signature objects — that focus is exactly what makes it powerful.

In short, Vidu bridges the gap between a fixed visual identity and limitless creative scenarios. It lets you take the subjects that matter most to your work and reliably reintroduce them in new poses, places, and stories, all guided by a simple written description and your choice of framing. Whether you're building a character universe, expanding a product campaign, or crafting a cohesive visual series, Vidu helps you stay consistent while staying creative.