Why creators choose Veo 3.1

Veo 3.1 is built for resolution. It reaches native 4K in a single generation, handles up to 8 seconds of video, and generates sound in the same pass. That gives you broadcast-ready UHD output without a separate upscale step. Veo 3.1 is Google DeepMind's flagship model for quality-first creation, built on the same architecture that powers Veo 3.1 Fast and Veo 3.1 Lite.

Four-K resolution in one pass

Veo 3.1 generates in 4K directly. No upscaling filter, no quality loss from stretching. You get the resolution the model understands natively.

Video that holds together across eight seconds

Eight seconds is long enough for an establishing shot plus a beat, or a product demo from two angles. The model keeps fine details consistent across the full duration, from fabric texture to water surface to light reflection.

References that anchor the frame

Start a clip from a reference image and Veo 3.1 moves from it, or use the image as direction for style and subject. The model takes that anchor into account throughout the generation.

Sound built into the same clip

Veo 3.1 can make sound together with the picture. Leave sound off for a score that you'll add, or get the ambient and the speech in one pass when they're baked into the concept.

How to use Veo 3.1 in Craze

You don't need to write video model prompts. You describe what you want, name Veo 3.1, and Craze makes it happen.

1. Start a project and describe the video or upload a reference image. If you have video you want to extend, add that clip.
2. Say you want Veo 3.1. Craze runs the model when you ask for it by name. If you don't name a model, Craze shows you compatible options with their costs.
3. Craze generates the clip at the resolution and with any sound you selected.
4. Review the result in the editor. Regenerate a shot that misses, or carry on editing.

Veo 3.1 on Craze also runs video extension, a 7-second continuation of a clip you provide. When you ask for it by name, Craze uses it for that shot.

Veo 3.1 capabilities and tasks

Veo 3.1 supports three generation tasks on Craze. Text-to-video takes a written prompt and generates a clip from scratch. Image-to-video takes a reference image and animates from it, preserving visual details and style. Video extension takes your existing clip and generates a continuation, keeping camera motion and framing intact. Sound generation runs in all three, producing dialogue, ambient audio, and effects timed to the visual action.

Veo 3.1 handles open-ended subjects. You can describe abstract concepts, specific objects, people, environments, or cinematography. The model writes to physical principles, so motion looks natural and objects behave according to real-world physics. Water flows downward, falling objects accelerate, and people walk with natural gait.

The model preserves detail across frames. Fine textures, fabric patterns, water surfaces, and lighting stay consistent from frame one through frame eight. Reference images act as anchors for composition and style, and Veo 3.1 holds that visual direction throughout the generation.

Veo 3.1 vs Veo 3.1 Fast

Both reach 4K and support 8-second clips with optional sound. Veo 3.1 Fast generates output faster. Veo 3.1 Fast costs less per second due to the faster generation. If speed is your priority or you're iterating through multiple takes, Veo 3.1 Fast is the practical choice. If you're shooting the hero shot and time doesn't matter, Veo 3.1 gives you the original quality.

Veo 3.1 pricing

Veo 3.1 is priced per second of video. Extension starts at $0.211 per second of added video; new clips use a separate rate. New clips can be 4, 6, or 8 seconds at 720p, 1080p, or 4K, with rates varying by resolution and sound setting. Extension adds a fixed 7 seconds at 720p.

Turning on sound increases the rate. Video extension tasks also use per-second billing. Use the cost calculator on this page to estimate your exact cost for your selected task, resolution, audio setting, and length.

Who Veo 3.1 is for

Commercial and broadcast work

Native 4K output means you hit production specs without an extra process. Motion graphics, product photography, and ads that will air on screens big and small all fit in one 8-second clip or a short sequence.

UHD portfolio and prestige projects

When you're shooting for native 4K, Veo 3.1 delivers it without upscaling. It's the standard for showreels and portfolio work where the visual quality is a statement, offering the full resolution production demands.

Commercials with sound on set

Veo 3.1 bundles sound in the same generation as the picture. When you need both to be part of the same creative decision, run them together.

Veo 3.1 vs other video models

Veo 3.1 is the resolution leader in the Veo family. Veo 3.1 Lite maxes out at 1080p and costs less per second, making it ideal for web and social delivery. Veo 3.1 reaches 4K and supports video extension, which Lite does not.

For a comprehensive comparison with other models on Craze, consider Gemini Omni Flash, which also generates 4K with built-in sound, or Grok Imagine 1.5, which focuses on image-to-video with strong character consistency. Veo 3.1 is the choice when you need native 4K resolution and the longest supported clip length.