Vidu Q3 Drama in the AI FILMS Studio image-to-video workspace

Open the video generation workspace and use the Select Generation Type dropdown to choose Image to Video. Then open the Select Model dropdown and pick Vidu Q3 Drama. The panel updates to show the script field, asset upload section, and the settings specific to Drama.

The model type selector lists all available image-to-video models. Choose Vidu Q3 Drama from the dropdown. The interface switches to the Drama layout with dedicated fields for script input and asset management.

Upload your primary reference image in the image source section. You can use an image from your device, a previous task output, or a public image URL. This image serves as the visual anchor for the scene.

Click Add Assets to define the cast of characters, locations, and props. Each asset gets a name, a type (character, scene, or tool), and an optional brief description. You can add between 1 and 14 assets per generation. Keeping the asset list between 3 and 10 entries typically produces the most consistent results. The model uses asset IDs to track and maintain each visual identity throughout the sequence.

The script field is the primary creative input. Write the full scene narrative in sequential sentences: character actions, emotional beats, camera direction, and lighting changes. The field accepts up to 5,000 characters. Clear, ordered writing produces more coherent motion than short or fragmented notes. Describe what happens at the start, what changes through the middle, and how the scene resolves.

The Video Style field pins the visual output to a specific aesthetic. Write a style descriptor of up to 30 characters, without punctuation. Examples: "3D modeling style", "2D animation style", "watercolor art style". Leave the field blank to use the default live action photorealistic output.

Duration sets the output clip length. The default range runs from 8 to 12 seconds. Begin with 8 seconds while testing a new script, then increase the duration once the structure and pacing are confirmed.

Aspect Ratio determines the output frame orientation. Choose 9:16 for vertical video intended for mobile or social platforms. Choose 16:9 for horizontal cinematic output suited to widescreen delivery.

Placement controls how reference images are used during scene construction. Upload is the default and fastest mode. Panorama synthesizes wide environments from a reference image, which works well for outdoor or architectural scenes that extend beyond a single frame. Step feeds frames from the previous scene forward into the next generation, creating a structured sequential flow across multiple outputs.

The workspace shows the exact credit cost for your selected duration and resolution before you submit. Confirm the settings, then click Create. The generation runs in the background and the completed video appears in your output panel when ready.

Vidu Q3 Turbo in the AI FILMS Studio image-to-video workspace

Go to the video generation workspace, set the generation type to Image to Video, and select Vidu Q3 Turbo from the model dropdown. Upload your reference image in the image source section before configuring the settings below.

Describe the motion, camera path, and scene changes you want in the prompt field. The model already knows what your image looks like from the upload. Focus the prompt on what changes over time: "a slow crane up revealing the city skyline behind the subject" gives the model specific direction to execute. The built in prompt enhancer refines your input before generation, so you do not need to add extensive technical syntax.

Movement Amplitude controls the overall intensity of motion in the output. Auto lets the model determine the level based on your image and prompt. Small produces subtle, refined motion suited to portraits and product photography. Medium delivers natural scene movement for most use cases. Large generates bold, dynamic animation for action sequences, cinematic reveals, or high energy scenes.

Vidu Q3 Turbo supports three resolution tiers. Use 540p to test composition and motion at the lowest credit cost. Move to 720p for standard deliverables and 1080p for final production output. Start at a lower resolution when iterating on a new prompt, then commit to 1080p once the motion reads correctly.

Set the clip length in the duration field. The default is 5 seconds, a good starting point for motion tests. Once the motion and composition are confirmed, increase the duration to generate a longer clip from the same settings.

Toggle Generate Audio to add synchronized environmental sounds to the clip. The model generates ambient audio based on the visual content: wind, rain, crowd noise, natural environments, and mechanical effects. This runs as part of the same generation and adds no extra credits.

Toggle Background Music to layer a harmonic score beneath the video. Both audio and background music can be enabled simultaneously. Neither adds to the credit cost for the generation.

Enter a number in the Seed field to lock the generation to a specific variation. The same image, prompt, and seed combination produces the same output on repeat runs, making it useful for A/B testing against a single motion baseline. Leave the field empty for a different random result each time. Click Create when your settings are ready.

Vidu Q3 Turbo image-to-video workflow in the AI FILMS Studio Nodes Graph Editor

Movement amplitude, resolution, duration, audio, and seed are all configurable inside the node panel. You can wire the video output directly into a Video Enhancer node to upscale the result, or route it to a Lipsync node to add synchronized dialogue in the same pipeline. For comparison, see the Kling 3.0 node pipeline tutorial, which follows the same connection pattern for image-to-video in the graph editor.