What Veo 3.1 Lite is

Veo 3.1 Lite is Google DeepMind's budget tier of Veo 3.1. On fal.ai you call fal-ai/veo3.1/lite for text-to-video and fal-ai/veo3.1/lite/image-to-video for image-to-video. The parameter surface is almost identical to the full Veo 3.1 endpoints, which means you can iterate on Lite and swap the endpoint string to promote a prompt to full quality later.

What it does well

Native audio generation is still here - dialogue, ambient, SFX - at 1/8 the cost. Same 8-second ceiling, same duration enum (4s, 6s, 8s), same lip-sync behavior. If you're drafting shots or running a product integration where cost-per-generation actually matters, Lite is where the Veo family becomes economically viable.

The tags on this model include lipsync and transform, which reflects what it's tuned for: animating characters and mouth movement from an image or prompt. Release date is recent - March 31, 2026 - and the schema matches that of full Veo 3.1 closely enough that a drop-in upgrade path actually works.

What it can't do

No 4K. Resolution is capped at 1080p, with 720p as default. That's the main tradeoff vs the full-tier model. Everything else - duration, aspect ratio options, audio, safety controls - matches the main endpoints.

Same structural limits as Veo 3.1: 8-second maximum, no end-frame parameter, aspect ratio is 16:9 or 9:16 only (plus auto on I2V). No multi-shot scripting like Kling's multi_prompt.

Parameters that matter

Image-to-video requires image_url at 720p or higher, in 16:9 or 9:16. Off-ratio images get cropped automatically.

Pricing

$0.05 per second flat. An 8-second 1080p clip is $0.40. A 4-second draft is $0.20. You can generate 8 Lite clips for the price of one full Veo 3.1 render at the same duration.

Working example

1import { fal } from "@fal-ai/client";
2
3const result = await fal.subscribe("fal-ai/veo3.1/lite", {
4 input: {
5 prompt: "A barista pulls an espresso shot, steam rising, 24mm lens, shallow depth of field",
6 duration: "6s",
7 resolution: "1080p",
8 aspect_ratio: "16:9",
9 generate_audio: true,
10 negative_prompt: "blurry, shaky camera, overexposed",
11 },
12});
13
14console.log(result.data.video.url);

When to pick Veo 3.1 Lite

Pick it for draft iteration on any Veo prompt, product integrations where per-generation cost is on the P&L, or any time 1080p is the ceiling you actually need. Skip it when 4K is a hard requirement (promote to full Veo 3.1). Skip it if you want clips longer than 8 seconds - both Veo tiers share that cap. For character animation with mouth movement at reasonable cost, Lite is the practical default in the Veo family.