← Home

Guide

Veo 3.1 on OpenClaw: Google Video for Your Agent

Veo 3.1 is Google DeepMind's current video model, and it is the one family in your agent's video dropdown you can reach two different ways — with a Google key or with a fal key. The tier you pick changes the price by more than 10x, and audio is billed separately. Here is how to choose.

Two Routes to the Same Model

Most video models in the OpenClaw Launch picker have exactly one provider. Veo has two:

  • Google — add a Google key and the Google group in your Video dropdown offers Veo 3.1, Veo 3.1 Fast and Veo 3.0 as curated rows. This is the direct route, and the same key powers Gemini for chat.
  • fal.ai — add a fal key and Veo 3.1 and Veo 3.1 Fast appear in the fal group instead, alongside Kling, Seedance and MiniMax H3. Useful if you would rather keep every video model on one bill.

Pick whichever key you already have. If you are running several video families, one fal key covering all of them is the simpler setup; if Veo is the only one you want, the Google route avoids adding a provider.

The Three Tiers and What They Cost

Veo bills per second of output, and audio is a separate line. These are the fal-side prices, which is where all three tiers are visible together:

Tier720p / 1080p, no audio720p / 1080p, with audio4K
Veo 3.1$0.20 / sec$0.40 / sec$0.40 no audio, $0.60 with
Veo 3.1 Fast$0.10 / sec$0.15 / sec$0.30 no audio, $0.35 with
Veo 3.1 Lite$0.03 (720p) / $0.05 (1080p)$0.05 (720p) / $0.08 (1080p)
Audio doubles the standard tier. A 5-second 1080p clip on Veo 3.1 with audio is $2.00. The same clip on Fast is $0.75, and on Lite it is $0.40. If your agent is answering casual requests rather than producing finished work, Lite or Fast is almost always the right default — and it is your own key being billed, not a plan credit.

Which Tier Should Your Agent Use?

  • Lite — the everyday default for a chat bot. At $0.05 per second for 720p with sound it is cheap enough to answer requests freely.
  • Fast — the middle ground when you want Veo's look without the standard-tier bill, and the tier Google exposes as a curated row on the direct route.
  • Veo 3.1 — reserve for output that ships. It is the most expensive per second of anything in the picker once audio and 4K are on.

Setting It Up

  1. Add either a Google key or a fal key on your API Keys page. Google keys come from Google AI Studio; fal keys from fal.ai/dashboard/keys.
  2. On your instance card, open the Video dropdown and select the Google or fal.ai provider group.
  3. Pick a Veo row. On the fal side the list refreshes daily from trending endpoints, so type “Veo” in the dropdown search box if the tier you want is not visible.
  4. Ask your agent for a clip in chat and the MP4 comes back in the conversation.

Model IDs for Self-Hosted OpenClaw

Google route:

google/veo-3.1-generate-preview
google/veo-3.1-fast-generate-preview
google/veo-3.0-generate-001

fal route (fal endpoint path with a fal/ prefix):

fal/fal-ai/veo3.1
fal/fal-ai/veo3.1/image-to-video
fal/fal-ai/veo3.1/fast
fal/fal-ai/veo3.1/fast/image-to-video
fal/fal-ai/veo3.1/lite/image-to-video

What's Next?

Give Your Agent a Video Model

Deploy an always-on agent, add a Google or fal key, and pick Veo 3.1.

Deploy Now