freycreate
By Google DeepMind·October 15, 2025

Veo 3.1

Google DeepMind (London / Mountain View, USA) — Google's flagship AI video model. Synchronized dialogue with precise lip-sync, frames-to-video for defined transitions — available as Standard and faster Fast variant.

Resolution1080p HD
Duration4 / 6 / 8 s
Audionative
Lip-syncprecise
VariantsStd / Fast
ReleaseOct 2025

Strengths

Precise dialogue + lip-sync

Dialogue is defined via quotation marks in the prompt. Veo 3.1 generates speech synchronized with lip movement — one of the few AI video models that pulls this off convincingly.

Frames-to-video

Define start and end frames as images and Veo generates the transition between them with audio. Perfect for precise storyboarding and continuity.

Long scenes up to 32 seconds

Chain multiple 8-second clips into continuous sequences with visual continuity. Character, lighting, and scene stay consistent. Currently up to 32 s (4 clips) at 720p / 1080p / 4K. — Coming soon in the UI.

Standard and Fast variant

Standard delivers top quality for final productions (~2-3 min per 8 s clip). The Fast variant generates 8 seconds in ~30-45 seconds — ideal for fast iteration.

Showcase

Real generations

Best for

Ad spots with dialogue and precise lip-sync

Storyboard-driven sequences (frames-to-video)

Cinematic brand content production

High-end social media reels

Not its strong suit

Native 1080p — upscale externally for 4K output (e.g. with our Upscaler Pro)

Standard variant takes 2-3 min per 8 s clip (Fast variant ~30-45 s)

Technical details

Architecture

Veo 3.1 is Google DeepMind's flagship video model. DeepMind operates from London (UK) and Mountain View (California). Veo 3.1 stands out for precise lip-sync in dialogue scenes and a scene-extension feature that stitches clips up to 148 seconds long.

Key features

  • Duration: 4, 6, or 8 seconds per generation
  • Native resolution: 1080p HD
  • Frames-to-video: defined start/end frames
  • Precise dialogue with lip-sync via quotation marks
  • Standard and Fast variant
  • Use from Claude Code, Cursor & Codex via freycreate MCP
Use directly from AI agents (MCP)

Ready to get started with Veo 3.1?

Beta access is free for selected creators. Claim your slot.

More models

Also interesting

We use cookies 🍪 ·