Google Veo 3 AI Video Generator: Veo 3.1 Fast, Pro and Lite
Veo 3 is Google DeepMind's family of video models, known for generating sound together with the picture. Zero Kit runs Veo 3.1, which replaces Veo 3, as Fast, Pro and Lite from one editor.
Veo 3 made sound part of the clip: instead of a silent video, it writes dialogue, sound effects and background ambience in the same pass as the image.
Veo 3.1 replaces it, with better image-to-video, closer adherence to the prompt and the option to set both the first and the last frame. That is why a search for Veo 3 or Veo3 ends up on Veo 3.1.
Google offers Veo 3.1 in Google Flow and on its Gemini developer platform.
Veo 3.1 comes in three versions:
- Veo 3.1 Fast is tuned for speed and everyday work such as ads, social clips and quick tests.
- Veo 3.1 Pro is the full-quality model, which Google lists simply as Veo 3.1.
- Veo 3.1 Lite is the lowest-cost version for drafts and high-volume work, and it stops short of the top resolution.
On Zero Kit the three share one page and one editor, on Plus and above. Fast returns every clip with sound, while Pro runs with sound off.
Every version turns a written scene or a photo into a clip of 4 / 6 / 8 seconds, landscape or vertical. Fast and Pro go up to 4k and also take a last frame, so the clip ends where you want.
Good Veo 3 prompts describe the sound as well as the picture: put spoken lines in quotation marks and name the effects and the ambience. The Veo 3.1 page has sample prompts and more tips.
To start from a still, the image to video AI page takes a photo and a description of the motion, and Animate photos makes a single picture move.
Video is charged per second of the clip, starting from 13 credits. The table below compares resolution, length, starting price and plan from the live model list.
Every Veo version compared
| Version | Resolutions | Length | Reference images | From | Plan |
|---|---|---|---|---|---|
| 720p, 1080p and 4k | 4 / 6 / 8 s | 2 | 13 credits / s | Plus |
Which Veo version should you use?
For clips with sound: Veo 3.1 Fast
Pick Veo 3.1 Fast when the clip should play with the sound on: a line of dialogue, a street or nature ambience, a product that clicks, pours or sizzles.
Every Fast clip on Zero Kit comes back with audio made together with the picture. It goes up to 4k and accepts a first and a last frame.
For the best picture: Veo 3.1 Pro
Choose Veo 3.1 Pro for the final take of a hero shot, where image quality matters more than the soundtrack. It is the full Veo 3.1 model, goes up to 4k and accepts a first and a last frame.
On Zero Kit it runs with sound off, so add music or a voiceover in the edit.
For drafts and volume: Veo 3.1 Lite
Use Veo 3.1 Lite to test ideas, compare several prompts or produce many short clips at the lowest rate.
It starts from text or a single photo, does not offer 4k and does not take a last frame. Once a prompt works, render the final version on Fast or Pro.
Veo versions on Zero Kit
Veo 3.1
Balances high cost-efficiency and performance, optimal for high-volume daily production.
Open Veo 3.1
Veo FAQ
Partly. The way to use Veo 3 for free is Google Flow, which gives every user free daily credits, enough for a few short Veo 3.1 Lite or Fast clips. Google AI subscriptions add more.
In the Gemini app, video is made with Gemini Omni rather than Veo.
On Zero Kit it is included in Plus and above, and each run costs credits. Signing up is free, so you can open the editor, write a prompt and see the price before you upgrade.
Veo 3.1 is the successor to Veo 3. Both generate sound together with the picture, and Veo 3.1 adds better image-to-video, closer prompt adherence and the option to set a last frame as well as a first.
Zero Kit runs Veo 3.1 in its Fast, Pro and Lite versions on Plus and above; the original Veo 3 is not offered here.
Yes. Veo 3 and Veo 3.1 generate dialogue, sound effects and background ambience together with the video.
On Zero Kit, where Veo 3.1 is included in Plus and above, Fast returns every clip with sound, while Pro runs with sound off, so pick Fast when the clip needs audio.
Yes. Veo 3.1 uses an uploaded photo as the first frame and keeps its subject, setting and light while adding motion.
Veo 3.1 Fast and Pro can also take a last frame, so the clip ends on a composition you choose. Veo 3.1 Lite takes a first frame only.
On Zero Kit, Veo 3.1 starts from 13 credits per second of video, and the rate depends on the version and the resolution. All three versions are included in Plus and above.
A clip is charged for its full length, and the Veo 3.1 page lists the live rate for every version and resolution.
Yes. Within the family, Veo 3.1 Lite and Fast cost less per second than the full model: Veo 3.1 Lite from 13 credits (about $0.13), Veo 3.1 Fast from 32 credits (about $0.32) and Veo 3.1 Pro from 60 credits (about $0.60).
Outside Google, Seedance 2.0 Fast and Seedance 2.0 Mini are lower-cost video models on Zero Kit, and each model's page lists its live per-second rates.
Dollar amounts are approximate, based on monthly subscription prices.
Veo 3, in its Veo 3.1 version. OpenAI discontinued the Sora app and removed its Sora video models from its developer platform, which makes Veo 3.1 the closest Sora alternative.
Veo 3.1 also makes short realistic clips with dialogue and sound effects made together with the picture, from a prompt or a photo.
Seedance 2.0 is strong at camera direction and Kling 3.0 at multi-shot clips.
Google offers Veo 3.1 in Google Flow and on its paid developer platforms. On Zero Kit it runs in your browser, on Plus and above:
- Create a Zero Kit account and open the editor.
- Pick Veo 3.1 Fast, Pro or Lite.
- Write a scene or upload a photo, choose a length and press generate.
No Google subscription, API key or code is needed, and you pay with Zero Kit credits.
Yes, on a paid plan. Images and videos made on a paid plan can be used commercially, including product pages, ads and social posts, subject to our Terms. Output from the free plan is for personal, non-commercial use only.
No. We do not use your prompts, uploads or results to train AI models. To run a generation, the prompt and any reference files are sent to the provider of the model you picked, which processes that request under its own terms.
Last updated: