How to Use Google Veo 3.1 in Google Flow, the Gemini App and the Gemini API (October 2026)

Quick answer: the place to use Veo 3.1 by name is Google Flow. Open a project, switch the prompt box from image to Video, pick Veo 3.1 Lite, Fast or Quality, and generate a 4, 6 or 8 second clip with sound. Developers can call the same model through the Gemini API on the paid tier. In the Gemini app, Google’s help page now describes video generation with Gemini Omni, so use Flow if you specifically want Veo 3.1. Everything here comes from Google’s help pages and docs, checked on October 18, 2026.

Where can you use Veo 3.1 in October 2026?

Surface	Veo 3.1 options	Who can use it	Official source
Google Flow (web, mobile app)	Veo 3.1 Lite, Fast and Quality, plus Gemini Omni Flash	18+, supported regions; Plus, Pro and Ultra add monthly credits	Flow Help
Gemini app	Video generation with Gemini Omni, per Google’s help page; Google’s plans page lists a limited Veo 3.1 Lite trial on Pro and Veo 3.1 on Ultra	Google AI plan required, 18+	Gemini Apps Help, Google AI plans
Gemini API	veo-3.1-generate-preview, veo-3.1-fast-generate-preview, veo-3.1-lite-generate-preview	Paid tier only	Veo docs, pricing

Google’s two Gemini app pages don’t say the same thing. If the Gemini app gives you an Omni video when you wanted Veo 3.1, switch to Flow, where you choose the model yourself.

What do you need before you start?

For Google Flow, Google’s getting-started page lists three requirements: you’re 18 or older with your age verified, you’re in a supported region, and you subscribe to Google AI Plus, Pro or Ultra. It adds that “VPNs will not enable access in unsupported regions”.

Google’s credits page describes a free allowance too. Everyone gets 50 Flow credits a day “with or without a subscription”. However, “during peak hours, roughly 2 PM to 5 PM UTC, you may not be able to generate videos without a subscription”. Some Workspace business and education plans get Flow with 50 daily credits at no extra charge.

Google’s two pages contradict each other. The getting-started page lists a Google AI Plus, Pro or Ultra subscription as a requirement, while the credits page says you get 50 credits a day “with or without a subscription”. If Flow asks you to subscribe, that’s why. Plan on a subscription if you need reliable video generation.

To compare Google’s plans with other AI subscriptions, see ChatGPT plans compared. For other tools with free tiers, see free AI video generators.

How do you make a Veo 3.1 video in Google Flow?

Google’s Create videos help page gives these steps:

1. Go to Google Flow. Google recommends a Chromium-based browser such as Chrome on a desktop.
2. Open an existing project or start a new one.
3. In the prompt box, describe the scene: subject, action, environment, lighting and style.
4. Click the model name. Flow defaults to the image model Nano Banana Pro, so click Video.
5. Optionally add Ingredients (reference images) or Frames (start and end images).
6. Choose the aspect ratio, number of outputs, the model (Veo 3.1 Lite, Fast or Quality) and the length.
7. Click Generate.

Flow also has an Agent mode. You turn it on in the prompt box and ask for things like several variations of one video at once.

Which Veo 3.1 tier should you pick?

Per Google’s models page and credits page, credit costs as of October 18, 2026:

Tier	Text to video	Frames (first, or first and last)	Ingredients (references)	Extend	Credits per generation
Veo 3.1 Lite	4, 6, 8 s	Yes	Yes (8 s only)	Yes (8 s clips)	10 (5 for Ultra)
Veo 3.1 Fast	4, 6, 8 s	Yes	Yes (8 s only)	No	20 (10 for Ultra)
Veo 3.1 Quality	4, 6, 8 s	Yes	No	No	100

Google’s credits page lists Quality for 8-second videos, while the models page lists 4, 6 and 8 seconds, so check the length options in Flow. Ultra subscribers can also run a lower-priority Veo 3.1 Lite mode for 0 credits. Google warns that “model costs are evolving fast”, so check the cost under Settings in the prompt box before you generate. You are not charged credits for failed generations.

How do you add a start and end frame?

Click the model name, then Video > Frames. Drag one image onto “+ Add start frame” and another onto “+ Add end frame”. Then describe the action or transition between them (Flow Help). The full technique is in first and last frame AI video.

How do you make a clip longer?

In Flow, “all Veo 3.1 8s videos can be extended, but you must use Veo 3.1 Lite to extend them” (Flow Help).

How do you use Veo 3.1 in the Gemini app?

Google’s Gemini Apps help page gives this flow:

1. Go to gemini.google.com and open the sidebar.
2. Click Create video. You can optionally pick a template.
3. Type your prompt. Add up to 5 images and one video if you want to build from your own media.
4. Submit. Generation can take a few minutes.
5. Under the video, click Share to download it or share it on YouTube.

You need a Google AI plan and must be 18 or older. The page describes this feature as running on Gemini Omni. Google’s plans page says Pro unlocks “a limited trial of Veo 3.1 Lite” and Ultra includes “video generation with Veo 3.1”.

How do you call Veo 3.1 from the Gemini API?

Call generate_videos with a Veo 3.1 model ID, poll the long-running operation until done, then download the file. This is shortened from Google’s Veo docs:

import time
from google import genai

client = genai.Client()  # reads GEMINI_API_KEY from the environment
op = client.models.generate_videos(
    model="veo-3.1-generate-preview",
    prompt="Drone shot following a red convertible along a coastal road at sunset",
)
while not op.done:
    time.sleep(10)
    op = client.operations.get(op)
video = op.response.generated_videos[0]
client.files.download(file=video.video, destination="veo_clip.mp4")

Things the docs flag:

• Paid tier only. Gemini API pricing lists Veo 3.1 as “Not available” on the free tier (pricing).
• Download within 2 days. “Generated videos are stored on the server for 2 days, after which they are removed.”
• Clip rules. 1080p, 4K and reference images require 8-second clips. Lite has no 4K and no reference images. Extension works on Veo 3.1 and Fast but not Lite, the opposite of Flow.
• Google’s default has moved. The Gemini API now recommends Gemini Omni Flash as its default video model (video overview).

To compare Google’s API with Runway, Kling and Luma, see AI video generation APIs compared.

How should you write a Veo 3.1 prompt?

Google’s prompt guide lists the elements to include: subject, action, style, and optionally camera positioning and motion, composition, focus and lens effects, and ambiance (Veo docs). For sound:

• put dialogue in quotes, e.g. "This must be the key," he murmured.
• describe sound effects explicitly, e.g. “tires screeching loudly”
• describe ambient noise, e.g. “a faint, eerie hum resonates in the background”

Google recommends English prompts for best results in Flow (Flow Help).

Does Veo 3.1 add a watermark?

Yes, an invisible one. “All outputs generated in Google Flow using Veo, Omni, or Nano Banana include invisible SynthID watermarks.” A separate “Visible watermarking” toggle sits under your profile picture. A visible watermark is applied automatically for users in India, South Korea and Vietnam who aren’t on the Google One AI Ultra 20x plan (Flow Help). For publishing rights, see Can you use AI videos commercially?

To see how Veo 3.1 compares with other models, read Veo 3.1 vs Kling vs Runway. All guides are in the AI video tools hub.