AI creators tools

Veo 3.1 video model

Name: Veo
Version: 3.1
Also Known As: Veo 3.1 Quality
Creator: Google DeepMind

Veo 3.1 dropped on October 16, 2025. Access is slowly rolling out to partners.This new version gives more control over how videos look and follow your story.

The native output resolution for Veo 3.1 is 720p or 1080p at 24 fps. Still only landscape 16x9 and portrait 9x16 options. The model can generate with and without audio, which the user can now select. Silemt videos are cheaper.

Video duration is also much more flexible with options for 4s, 6s, 8s

It gets your story better now and captures more realistic   textures. It also turns images into videos with better quality and smoother transitions.

Use images to shape scenes. Drop in multiple pics with people or stuff and Veo will put them into one scene with sound (video ingredients).

Longer scenes. You can now make clips that go past 8 seconds, even up to a minute. It keeps the same people and background by using the final second of the last clip to keep the story going.

Start-end frames. Pick how your scene starts and ends. Just set the first and last frames and Veo fills in the middle.

More details, more fun. You can add anything from small objects to fantasy creatures. It blends them in by matching shadows and lighting so nothing looks off.

Better sound and feel. Audio’s richer, voices feel more natural, and the videos come out sharper. Prompts now match better too.

Creative tools now work with audio. These features now have audio support:

  • Ingredients to Video. Use pics to guide the style and keep things consistent.
  • First and Last Frame. Lock in where your story starts and stops.
  • Extend. Build on past clips to keep your story going.

Veo 3.1 Examples

Overall, this model does rather well with this prompt. Generated on October 19, 2025
Compare Models
Nice, realistic text to video with good motion. Could've been quicker but it's nice. Generated on October 16, 2025
Compare Models
That's a really nice, detailed, realistic image-to-video generation. Generated on October 16, 2025
Compare Models
This image to video prompt worked nicely. Although no 'arc shot' and there's a fish awkwardly materializing out of nowhere. Generated on October 16, 2025
Timelapse is great in Veo 3.1 Generated on October 16, 2025
Compare Models
Follows prompt, anatomically sound but boring slow movements Generated on October 16, 2025
Compare Models
Text to video example of using Veo 3.1 model Generated on October 15, 2025
for link to original.

Where To Find Veo 3.1

If you'd like to access this model, you can explore the following possibilities:

Other Models by Google DeepMind