Google DeepMind's video model makes 4, 6 or 8-second clips with dialogue, music and sound effects — up to 4K, from text, an image, first and last frames, or three reference images.
Dialogue
Noir dialogue
The older detective lights a cigarette and says: "She was here tonight." The younger one replies: "Then we're already too late." Rain patters, distant siren
Quick answer
Veo 3.1 is Google DeepMind's video model. Veo 3 (Google I/O, 20 May 2025) added native audio; Veo 3.1 (15 October 2025) added reference images, first-and-last-frame control and extend. It makes 4, 6 or 8-second clips at 720p, 1080p or 4K, with dialogue, sound effects and music generated in the same pass. In Kenerate's studio you can run Veo 3.1, Veo 3.1 Fast, Veo 3 and Veo 3 Fast — text or image to video, up to 3 reference images, and extend by 7 seconds at a time. There is no Veo 4; Google's newest video model is Gemini Omni Flash (May 2026), also in the studio. The clips here are real Veo 3.1 outputs.
Updated 27 September 2026 · specs read from Kenerate's code
Real demo
Switch between the two and turn the sound on. Same start image and motion prompt, two models.
Both are unedited outputs from Kenerate's video studio.
Try Veo 3.1Short answers
Google DeepMind's video model: 4, 6 or 8-second clips at up to 4K with dialogue, sound effects and music, from text, an image, first and last frames or up to 3 reference images.
In Google's Gemini app, video needs a Google AI Plus, Pro or Ultra plan; YouTube Shorts offers a free Veo 3 Fast in some countries. On Kenerate, Veo runs on one-time credit packs that never expire.
Pick Veo 3.1 in the video studio, start from text or an image, write the prompt with any dialogue in quotes, choose 4, 6 or 8 seconds and a resolution, and generate.
4, 6 or 8 seconds per clip. Extend adds 7 seconds at a time.
Real clips
Five Veo 3.1 image-to-video clips (6 s, 720p, with sound) — dialogue, music, kitchens, storms and ice.
Real Veo 3.1 clips (6 s, 720p, with sound). Turn the sound on — Veo generates dialogue and effects.
How it works
Start image, a prompt with the dialogue in quotes, Veo 3.1.
01
Seedream 5.0 Pro at 2K — or upload your own photo.
02
The older detective lights a cigarette and says: "She was here tonight." The younger one replies: "Then we're already too late." Rain patters, distant siren
03
Veo 3.1 · 6 s · 720p · with sound. The clip on the left is the real result.
Tips
From Google's Veo prompt guide.
Step by step
Engines
From the studio's own model list.
Google DeepMind
Veo 3.1Newest Veo
Dialogue and cinematic sound
Google DeepMind
Veo 3.1 Fast
Faster drafts
Google DeepMind
Veo 3
The original sound model
Google DeepMind
Veo 3 Fast
Quick Veo 3 clips
Gemini Omni 1.1 Flash
Editing a video by describing the change
The clips above were made with Veo 3.1; the comparison uses Wan 3.
Every clip on this page started as one still image. Make yours, then animate it.
In depth
What the models can do, how to prompt them, and what isn't allowed.
Veo 3 (May 2025) was the first Veo with native audio. Veo 3.1 (October 2025) added up to three reference images ('ingredients'), first-and-last-frame interpolation and extend; a January 2026 update added native 9:16 and 4K.
Google ships Veo in the Gemini app, Flow, YouTube, Google Vids, the Gemini API and Vertex AI. In Kenerate it sits next to Seedance, Kling, Wan and 50+ other models.
FAQ
7 questions
Google DeepMind's video model: 4, 6 or 8-second clips at up to 4K with dialogue, sound effects and music, from text, an image, first and last frames or up to 3 reference images.
In Google's Gemini app, video needs a Google AI Plus, Pro or Ultra plan; YouTube Shorts offers a free Veo 3 Fast in some countries. On Kenerate, Veo runs on one-time credit packs that never expire.
Pick Veo 3.1 in the video studio, start from text or an image, write the prompt with any dialogue in quotes, choose 4, 6 or 8 seconds and a resolution, and generate.
4, 6 or 8 seconds per clip. Extend adds 7 seconds at a time.
Yes — Veo 3 and 3.1 generate dialogue, sound effects and ambient audio with the video.
No. Google's newest video model is Gemini Omni Flash (May 2026), which you can also run in Kenerate's studio.
No. Google's generative AI prohibited use policy bans sexually explicit content.
Explore
Generators, tools, models and guides — one studio, one credit balance.
Written by the Kenerate AI team
We build and run Kenerate's video studio. Every clip here is a real, unedited Wan 3 output (5 s, 720p) and every start image a real Seedream 5.0 Pro output; model specs and limits were read from the app's code; facts about other tools link to their own pages. Last updated 27 September 2026. Corrections: hello@kenerateai.com.
Ready when you are
One studio for images and video.
