01
Is Veo 3.1 free here?
No. Veo 3.1 and Veo 3.1 Fast are paid-plan models, and each generation is metered in credits. If you want to test the platform without paying first, the free plan includes Seedance 2.0 Mini.
Veo 3.1 online:text and image to video in one workspace
Write one shot, choose 4, 6 or 8 seconds, and let Veo 3.1 generate it with synchronized dialogue and sound effects — then keep the result beside every Seedance generation. Start in the composer below.
Specifications
What Veo 3.1 does here
The limits below are the ones this generator actually applies, not the vendor headline numbers. Where your plan sets a lower ceiling, the row says so.
| Models in the menu | Veo 3.1 for quality and Veo 3.1 Fast for cheaper, faster generations. Both sit under the Google filter in the model menu. |
|---|---|
| Clip length | 4, 6 or 8 seconds per generation. |
| Output resolution | Veo 3.1 offers 720p and 1080p; Veo 3.1 Fast adds 4K. Your plan caps the ceiling: 1080p from Lite, 2K on Pro, 4K on Premium. |
| Aspect ratios | 16:9 landscape and 9:16 vertical, generated natively rather than cropped from a wider frame. |
| Image input | Attach one image as the start frame, or two images to set start and end frames and let the model direct the transition. |
| Audio | Veo generates synchronized dialogue and sound effects itself, so there is no separate audio switch in the panel. |
| Prompt length | Up to 2,000 characters per prompt. |
| Plan requirement | Both Veo variants need a paid plan. The free plan covers the entry Seedance models instead. |
| Watermark | Google embeds an invisible SynthID watermark in Veo output. That is the model vendor’s behaviour, not a platform overlay. |
Official Veo sources
The background clip on this page was generated on SeedVideo AI with Seedance, not with Veo 3.1.
Last updated:
Straight answers
Before you spend a credit
The four questions people ask first when they land on a Veo 3.1 page that is not run by Google.
02
Veo 3.1 or Veo 3.1 Fast?
Fast costs less per second and is the only one of the two that can output 4K. Veo 3.1 is the quality-first choice. Prompts, durations and frame inputs are identical, so you can switch between them mid-project.
03
Can I animate a single image?
Yes. One attached image becomes the start frame. Attach a second and the model generates the transition from the first frame to the last instead of extending freely.
04
What does one clip cost?
Credits are charged per second, so an 8-second clip costs twice a 4-second one at the same settings. The composer shows the exact quote before you submit, and nothing is deducted for a failed task.
Workflow
From prompt to Veo clip
The same four steps whether you start from text, one image, or a pair of frames.
- 01
Select Veo 3.1
Open the composer on this page, or the full generator, and pick Veo 3.1 or Veo 3.1 Fast from the model menu under the Google filter.
- 02
Write one shot
Name the subject, the setting, the action, the camera move and the last beat. Add the dialogue or sound you want, since Veo generates its own audio.
- 03
Attach frames if needed
Upload a start frame to animate an existing image, or a start and end frame when the shot has to begin and finish on specific visuals.
- 04
Set length, then submit
Choose 4, 6 or 8 seconds and the resolution your plan allows, check the credit quote, and submit. Finished clips land in your asset library.
Comparison
Veo 3.1 vs Seedance 2.5
Both models run in this workspace, so the choice is per shot rather than per subscription. Length and reference depth are where they separate.
| Feature | Veo 3.1 | Seedance 2.5 |
|---|---|---|
| Clip length | 4, 6 or 8 seconds | Up to 30 seconds in a single generation |
| Shot structure | One continuous shot per generation | Multi-shot sequences from one prompt |
| References | Start frame, or start plus end frame | Up to 30 reference images and 10 reference videos |
| Resolution | 720p and 1080p; 4K on Veo 3.1 Fast | 480p, 720p and 1080p |
| Audio | Always generated with the video | Optional — audio generation is a switch |
| Editing the result | Regenerate with a new prompt | Localized edit and extend on an existing video |
| Free plan | Not included | Not included — Seedance 2.0 Mini is the free-plan model |
Vendor reference: Google DeepMind Veo model page
Where it fits
Shots Veo 3.1 handles well
Short, dense, audio-led shots are its strength; long multi-shot narratives are not.
01
Audio-led product beats
An eight-second hero shot where the sound design carries as much as the frame: a click, a pour, a close-up with room tone.
02
Two-frame transitions
You have the opening still and the closing still. Set both as frames and let the model invent the motion between them.
03
Vertical social openers
Native 9:16 generation for the first three seconds of a short-form ad, without cropping a landscape render.
04
Dialogue tests
Draft a spoken line in-scene to hear whether the delivery and lip timing work before booking a shoot or a voice session.
Also on this platform
If Veo 3.1 is not the right fit
Same account, same credits, same asset library. These three Seedance pages cover the cases Veo 3.1 cannot take.
01
Testing on the free plan
Veo needs a paid plan. Seedance 2.0 Mini is the entry model included on the free plan, at 480p and 720p, for cheap iteration before you commit.
Try Seedance 2.0 Mini free online02
Shots longer than 8 seconds
Veo caps a generation at 8 seconds. Seedance 2.5 runs to 30 seconds with multi-shot structure and up to 30 reference images in one prompt.
See Seedance 2.5 for 30-second video03
Reference-heavy production
When a product, character or style has to stay identical across takes, the Seedance 2.0 reference workflow holds it better than a single start frame.
Compare the Seedance 2.0 video generatorFAQ
Veo 3.1 questions
Model scope, plan requirements, watermarking and how this platform relates to Google.
Veo 3.1 is Google DeepMind’s video generation model, released through the Gemini API in preview alongside a cheaper Veo 3.1 Fast variant. It generates short clips with synchronized native audio and supports first-and-last-frame direction.
Veo 3.1 and Veo 3.1 Fast. Both accept text prompts, a start frame, or a start and end frame, at 4, 6 or 8 seconds, in 16:9 or 9:16.
Yes. Choose 9:16 and the frame is generated vertically rather than cropped from a landscape render, which keeps faces and product labels inside the safe area.
Yes. Audio is part of what the model generates, including dialogue and sound effects timed to the action, so there is no separate audio toggle for Veo in the panel.
Switch models rather than stitching. Seedance 2.5 generates up to 30 seconds in one pass with multi-shot structure, and it accepts far more reference material per prompt.
No. SeedVideo AI is an independent platform that runs Veo 3.1 through an upstream API, and it is not operated by, endorsed by or affiliated with Google. The official Veo links on this page point to Google’s own documentation.
Veo 3.1 on SeedVideo AI
Generate your first Veo shot
Prompt, duration, frames, submit. Every generation lands in the same asset library as your Seedance work.
Open the generator