Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.imagine.art/llms.txt

Use this file to discover all available pages before exploring further.

VIDEO MODELby Google DeepMindVeo 3.1 family

Google Veo 3.1 Fast

Google DeepMind’s balanced Veo 3.1 variant — up to 4K resolution, 4, 6, or 8-second clips with 3 reference image support, and faster generation times than the flagship Veo 3.1 for production workflows.

Resolution
Up to 4K
Duration
4, 6, or 8 seconds
Audio
No
References
Up to 3 images

Balanced speed and quality

Veo 3.1 Fast sits in the middle of the Veo 3.1 family — faster than the flagship Veo 3.1 with more capability than Veo 3.1 Lite. Multi-reference image input (up to 3 images) and 4K resolution support are included. Generation times are 60–90 seconds for 720p and 90–120 seconds for 1080p, making it practical for production workflows where quality and speed need to be balanced. The Transformer backbone with spatio-temporal patches is shared across the Veo 3.1 family.

Capabilities

Up to 4K resolution

Supports 720p, 1080p, and 4K output — choose the resolution tier that fits your delivery requirements.

3 reference images

Multi-reference input with up to 3 images for subject appearance, visual style, and scene composition anchoring.

4, 6, or 8-second clips

Selectable clip length — choose 4, 6, or 8 seconds depending on content needs and credit budget.

Frame-to-frame generation

Supports image-to-video with natural, physically plausible motion from a reference starting frame.

Faster than flagship Veo 3.1

Shorter generation times than Veo 3.1 — 60–120 seconds at 720p–1080p for production-pace workflows.

Veo 3.1 family comparison

ModelAudioDurationMax resSpeedCost
Veo 3.1 LiteNo8s1080pFastLowest
Veo 3.1 FastYes4/6/8s4KBalancedMedium
Veo 3.1Yes4/6/8s4KSlowerHighest

Specifications

FeatureDetails
DeveloperGoogle DeepMind
Resolution720p, 1080p, 4K
Duration4, 6, or 8 seconds (selectable)
Frame rate24 FPS
AudioNo native audio
Reference imagesUp to 3
Aspect ratios16:9, 9:16
Generation time~60–90s (720p), ~90–120s (1080p), ~2–3min (4K)
ArchitectureTransformer backbone, spatio-temporal patches

How to use

1

Open the AI Video Generator

Log into ImagineArt and go to the AI Video Generator.
2

Select Google Veo 3.1 Fast

Choose Google Veo 3.1 Fast from the model dropdown.
3

Write your prompt

Include scene description, subject behavior, camera movement, and audio environment details.
4

Upload reference images (optional)

Add up to 3 reference images for character appearance or visual style anchoring.
5

Select resolution

Choose 720p, 1080p, or 4K depending on your output requirements and credit budget.
6

Generate

Click Generate and receive your output.

Prompting tips

  • Describe the visual scene in detail — “A waterfall cascades in the background, mist rising into the air” sets a vivid visual scene.
  • Use reference images for product or character consistency — Upload a product shot or character photo as a reference to anchor the visual in your generated clip.
  • Be specific about camera framing — “Tight close-up,” “wide establishing shot,” or “over-the-shoulder angle” guide Veo 3.1 Fast’s framing decisions.

Example prompts

A barista steams milk in an artisan coffee shop. Close-up on the steam wand, foam forming. Warm, cozy lighting. 8 seconds, 1080p.
A coastal drone shot at sunrise. Wide angle, slow forward movement over calm ocean. Golden light, misty horizon. 8 seconds, 4K.

Compare models

ModelAudioDurationResolutionSpeedBest for
Veo 3.1 FastYes4/6/8sUp to 4KBalancedAudio-visual production, 4K
Veo 3.1 LiteNo8s1080pFastestCost-efficient, no audio
Veo 3.1Yes4–8sUp to 4KSlowestMax fidelity, broadcast quality
Sora 2 ProYes25s1080pStandardLong-form A/V, physics
Veo 3.1 Fast is the practical default choice in the Veo 3.1 family — it includes audio, supports 4K, and generates faster than the flagship. Move up to Veo 3.1 when you need the highest visual fidelity and full audio including dialogue and voiceover.