Video model · by Google DeepMind
Make short product videos with sound using Google's Veo 3.1. Control the first and last frame, or lock the look with reference images, in the same workbench as every other model.
Veo 3.1 is Google DeepMind's video generation model, released in October 2025 as an upgrade to Veo 3, with audio generated alongside the picture. On Avocado AI you get Veo 3.1, Veo 3.1 Fast and Veo 3.1 Reference for 4 to 8 second clips at up to 1080p.
For ecommerce brands
Set a start frame and an end frame and let Veo 3.1 animate the move between them.
Dialogue, ambience and effects generated with the video, ready for Reels and TikTok.
Veo 3.1 Reference uses up to 3 reference images to hold a product or style.
Specs on Avocado
| Made by | Google DeepMind |
|---|---|
| Versions on Avocado | Veo 3.1, Veo 3.1 Fast, Veo 3.1 Reference |
| Modes on Avocado | Text to video, image to video, start and end frames, reference |
| Durations | 4, 6 or 8 seconds (Reference: 8 seconds) |
| Resolutions | 720p or 1080p (Fast: 720p) |
| Aspect ratios | 16:9, 9:16, 1:1 (start and end frames: 16:9, 9:16) |
| Credits per clip | Veo 3.1: 48 with audio, 30 without. Fast: 25 / 18. Reference: 50 |
Credits are the live workbench prices. The Generate button always shows the exact cost before you create. Plans in EUR, see pricing.
Veo 3.1 suits short product reveals, unboxing moments and ads that need sound. Start from your product photo, or give a first and last frame so the move lands exactly where you want. Veo 3.1 Reference takes up to 3 reference images to keep a product or style consistent.
Veo 3.1 is priced per clip. A clip costs 48 credits with audio or 30 without. Veo 3.1 Fast costs 25 with audio or 18 without, and Veo 3.1 Reference costs 50. Plans start at EUR 19.99 a month for 100 credits, and the exact cost shows before you generate.
You upload two images: the first frame and the last frame of the clip. Veo 3.1 generates the motion between them, which is useful for a product turning to face the camera or a pack opening. In this mode the output is 16:9 or 9:16.
Use Veo 3.1 Fast to test ideas: it renders at 720p and costs 25 credits with audio. Switch to Veo 3.1 for the final version when you want 1080p and the most detail, at 48 credits with audio. Both support start and end frames.
Updated September 26, 2026
No. Veo 3.1 is Google's upgrade to Veo 3, with better audio and picture sync and more creative controls. Avocado offers the Veo 3.1 family: Veo 3.1, Veo 3.1 Fast and Veo 3.1 Reference. The older Veo 3 models are no longer in the video workbench.
Veo 3.1 and Veo 3.1 Fast make 4, 6 or 8 second clips, and Veo 3.1 Reference makes 8 second clips. For a longer ad, join several clips in a Flow with Director mode, or pick Seedance 2.5, which runs up to 30 seconds in one clip.
Yes. Audio is on by default, and you can switch it off in the video workbench. A Veo 3.1 clip without audio costs 30 credits instead of 48. That suits videos where you plan to add your own music or a voiceover from Audio Studio.
Yes. Every paid plan includes full commercial rights and no watermark, so you can run what you make in paid ads, on product pages, in email and on social. Results are saved to your Avocado library, ready to reuse in a video, a Flow or a Super Agent post.
Veo 3.1 on Avocado supports 16:9, 9:16 and 1:1. When you use start and end frames, the options are 16:9 and 9:16. Pick 9:16 for Reels, TikTok and Stories, 1:1 for feed posts and 16:9 for YouTube and product pages.
Veo 3.1 is precise for short clips with start and end frames. Seedance 2.5 runs longer, up to 30 seconds, and costs less per short clip at 480p or 720p. You can test both on the same product photo in the video workbench and keep whichever looks right.
More models
Google's fast image model for product shots and edits, up to 4K.
ByteDance's video model for product ads and UGC, 4 to 30 seconds with native audio.
Kuaishou's video model in Standard, Pro and native 4K, 3 to 15 seconds.
OpenAI's image model and Avocado's default, for product images and precise edits.
Veo 3.1 and every other image and video model in one workspace, on one credit balance.
