Generate video with Google Veo 3.1
Google DeepMind's cinematic video model — with audio baked in
Veo 3.1 is Google DeepMind's premium video model, delivered on Singular Lab through Google's Vertex AI. It's the model to reach for when the shot needs to feel like a real film: cinematic lensing, controlled lighting, and — uniquely — an audio track (dialogue, ambient, music) generated together with the picture so the two never drift out of sync.
What Veo 3.1 is best at
Audio-native generation — dialogue + ambient + music baked into the video
Cinematic camera language: dolly, crane, whip-pan handled correctly
Physical light + shadow behaves like a real cinematographer's rig
First-frame image reference for continuity across shots
Fast tier (Veo 3.1 Fast) for iteration; standard tier for finals
8-second clips in 16:9 and 9:16, 720p / 1080p
Who's using Veo 3.1 on Singular Lab
- Short-form ad teams needing on-brand audio and video in one pass
- Music-video creators who want the score baked in, not mixed later
- Film-school and pre-vis workflows where cinematic language matters
- Anyone who kept switching between video-model and TTS to sync audio
FAQ
Do I need a Google Cloud account for Veo 3.1?
+
Veo 3.1 vs Veo 3.1 Fast — what's the difference?
+
Does Veo 3.1 really include audio?
+
How does Veo 3.1 compare to Sora 2?
+
Try Google Veo 3.1 free
Opens Manual Creation with Google Veo 3.1 preselected.