Official Google Vertex AI
# Veo 3.1 (G) High-fidelity cinematic video generation for professional storytelling. ## Overview Veo 3.1 (G) represents the pinnacle of Google’s generative video technology, specifically engineered for high-end creative production and marketing workflows. By leveraging advanced diffusion architectures, this model excels at maintaining temporal consistency, realistic motion physics, and high-fidelity visual textures. Unlike standard video models, Veo 3.1 is optimized for professional cinematic standards, ensuring that lighting, camera movement, and subject interaction feel intentional and polished. This model is distinct in its ability to handle complex prompts with nuanced cinematic language. Whether you are looking to generate sweeping landscape shots, detailed product close-ups, or character-driven sequences, Veo 3.1 interprets creative intent with precision. Its integration with Gemini-powered prompt enhancement ensures that even concise user inputs are expanded into rich, descriptive instructions that the model can translate into high-quality visual assets. ## Use Cases * **Brand Storytelling:** Generate high-impact B-roll footage for commercial spots without the need for expensive location shoots. * **Product Visualization:** Create cinematic product reveal videos by utilizing reference images to maintain brand consistency. * **Social Media Campaigns:** Produce platform-specific content in 9:16 aspect ratio for TikTok, Instagram Reels, and YouTube Shorts. * **Storyboarding & Pre-visualization:** Quickly iterate on visual concepts for ad campaigns to align creative teams before full production. * **Dynamic Transitions:** Use the video transition capabilities to create seamless, high-production-value cuts between scene segments. * **Asset Generation:** Generate custom background plates or atmospheric video elements for motion graphics and compositing. ## Parameters Veo 3.1 (G) offers a comprehensive suite of controls to fine-tune your output. Please refer to the Input Parameters section in the Pixloop interface to adjust settings such as resolution, aspect ratio, duration, and person generation constraints to match your project requirements. ## Tips for Best Results * **Leverage Prompt Enhancement:** Keep the `enhance_prompt` toggle enabled to allow Gemini to add descriptive cinematic terminology that improves lighting and camera movement quality. * **Use Reference Images:** For consistent product or brand identity, upload reference images to guide the model's visual style and subject matter. * **Iterate with Seeds:** If you find a composition you like but want to adjust the motion, keep the seed consistent while tweaking the prompt to maintain visual continuity. * **Mind the Duration:** For high-action sequences, use shorter durations (4s) to ensure the model maintains focus, while reserving 8s for slower, more atmospheric shots. * **Negative Prompting:** Use the `negative_prompt` field to explicitly exclude unwanted elements like "blurry," "distorted faces," or "low resolution" to clean up your output. ## About Veo 3.1 (G) is a state-of-the-art video generation model developed by Google, integrated into the Vertex AI ecosystem. It builds upon the lineage of Google's research into large-scale generative models, focusing on high-resolution, temporally stable video synthesis. By combining deep understanding of visual semantics with advanced motion modeling, it provides marketing teams with a robust tool for professional-grade creative asset generation.
Model input reference derived from preset schema.
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
| seed | number | No | null | Seed for deterministic video generation (0-4,294,967,295) |
| image | image | No | null | Base image for image-to-video generation. Supports JPEG, PNG formats |
| prompt | string | Yes | null | Text prompt describing the video you want to generate |
| last_image | image | No | null | Last frame image for video generation (optional) |
| resolution | select | No | "720p" | Resolution of the generated video |
| first_image | image | No | null | First frame image for video generation (optional) |
| storage_uri | string | No | null | Custom GCS storage location for output videos |
| aspect_ratio | select | No | "16:9" | Aspect ratio of the generated video |
| sample_count | select | No | "1" | Number of video variations to generate (1-4) |
| enhance_prompt | boolean | No | true | Use AI to enhance the prompt quality using Gemini |
| generate_audio | boolean | Yes | true | Whether to generate audio for the video (required for Veo 3.1 models) |
| negative_prompt | string | No | null | Describe content you want to prevent in the video |
| duration_seconds | select | No | "8" | Duration of the generated video in seconds (4, 6, or 8) |
| reference_images | image | No | [] | Reference images for style or asset guidance (max 3 images) |
| person_generation | select | No | "allow_adult" | Control people/face generation in videos |
| compression_quality | select | No | "optimized" | Video compression quality |