Google

Veo 3.1 Reference-to-video pricing

google/veo3.1/reference-to-video

Veo 3.1 keeps a subject's look from up to three photos, in an 8 second clip.

From, per second
$0.20
✦01 · Pricing

What it costs

Per second, and what a single clip comes to at each length. These are computed rather than multiplied out, so the minimum billed unit and any cap are already applied.

Tierper second8s clip
720p$0.20$1.60
1080p$0.20$1.60
4k$0.40$3.20
720p with audio$0.40$3.20
1080p with audio$0.40$3.20
4k with audio$0.60$4.80
Duration
8 seconds, set by the model
Billing
per clip, whatever its settings
✦02 · Capabilities

What it accepts

Aspect ratio

Framing is set by the model.

Resolution tiers

Priced per tier.

Audio

Bills at its own higher rate.

Negative prompt

Accepts a negative prompt.

Reference photos

Takes up to 3 photos of people, products and places.

Fixed durations

Any whole number of seconds in range.

✦04 · Alternatives

Others in this lane

Photos of who and what is in the shot steer a new clip, and the people and products keep their look. Billed per output second, so the length is the dominant term in the bill and the resolution tier is usually the second.

ModelFrom, per second
PixVerse V6$0.025
Gemini Omni 1.1 Flash$0.03
PixVerse C1$0.03
Vidu Q4$0.045
Wan 3.0$0.05
MiniMax H3 Open Weights$0.05
HappyHorse 1.1$0.07
Vidu Q3$0.07
Wan 3.0 Prime$0.075
Veo 3.1 Fast$0.08
Grok Imagine Video 1.5$0.082
Kling O3 Std$0.084
MiniMax H3$0.10
SkyReels V4$0.11
Kling O3 Pro$0.112