HeyOz HeyOz
VEO 3.1 BY GOOGLE DEEPMIND

Veo 3.1

Generate cinematic video with audio produced in the same pass — dialogue, sound effects, and ambience together. Guide a shot with up to three reference images, choose 4, 6, or 8 seconds, and render at 720p, 1080p, or up to 4K.

What people are generating

Spots, scenes, and product moments made with Veo 3.1 — audio included, straight out of the generation.

Start creating for free

How To Use The Veo 3.1 AI Video Generator

Turn a brief into a finished shot with Google Veo 3.1 in three steps.

PICK A MODEL
STEP 1

PICK A MODEL

Select Veo 3.1 as your model. Choose the standard tier for final quality, or Fast when you're testing angles.

PROMPT/ADD IMAGE
STEP 2

PROMPT/ADD IMAGE

Describe the shot, the audio you want, and the camera move. Add up to three reference images to guide a character, a product, a set, or a look — or give a first and last frame and let the model fill the middle.

GENERATE AND ADJUST
STEP 3

GENERATE AND ADJUST

Watch it back, tune the prompt, and regenerate. When a clip is right, extend it to carry the story past the eight-second mark.

Key Features of Veo 3.1

Veo 3.1 on HeyOz gives you reference-guided shots, audio generated alongside the picture, and clips you can extend into a sequence — wired into your brand kit and your ad formats.

INGREDIENTS TO VIDEO

Reference Images That Guide The Shot

Upload up to three images — a spokesperson, your product, a set, a style plate — and Veo 3.1 references them while it generates. Google added this in 3.1; earlier Veo generations had no reference-image path at all. It's how a mascot or a real SKU stays recognisable from one shot to the next instead of drifting every time you re-roll.

Try it Now
SCENE EXTENSION

Extend Past One Generation

Each generation runs 8 seconds at most. Scene extension continues a clip you already have, so you can chain beats into a longer narrative rather than stitching unrelated takes. Two things to plan for: extension renders at 720p, so finish your hero shot at the tier you intend to ship, and the Lite tier doesn't extend.

Try it Now
NATIVE AUDIO

Sound Generated With The Picture

Dialogue with matching lip movement, sound effects, and room tone arrive with the video in one pass — no separate scoring step. Veo 3.1 is Google's richer take on the audio the family has carried since Veo 3. Worth knowing: the Add and Remove Object edit still runs on Veo 2 and produces no audio, so treat it as a picture-only tool.

Try it Now
FORMATS & DURATION

Cut To Length, Sized For The Placement

Pick 4, 6, or 8 seconds — the older Veo generations only ever gave you 8. Render at 16:9, 24fps, for YouTube and display placements. Resolution runs 720p, 1080p, and up to 4K. Note that the longer 8-second setting is required once you move to the higher resolutions or use the reference features.

Try it Now

Professional videos for every use case

Hook-length social cuts, TV-style spots, and multi-shot campaigns — with the sound already in them.

Direct-Response And CTV Spots

Google points to work from QuickFrame AI and a partnership with MNTN, both using Veo to build TV and digital ads without a traditional production cycle. Those are Google's own partner accounts, not independent tests — but the shape of the workflow is real, and it's the one this model is tuned for.

Social Ad Variants

Hook-length cuts at 4, 6, or 8 seconds, sized for the feed. That maps cleanly onto the short social placements where most testing happens.

Multi-Shot Brand Campaigns

Give the model your spokesperson, mascot, or packshot as references and build a run of shots that read as the same campaign — then extend the ones that land.

Product Transformations

Set a first frame and a last frame and let Veo 3.1 generate the move between them, audio included. Google flags this path for controlled camera moves and before-after reveals.

Volume Testing

Google cites Pocket FM running Veo across 100+ team members weekly and reporting a 30–40% uplift in user retention against live-action promos. It's a vendor-supplied case study, so read it as a direction rather than a benchmark — but volume creative is what the Fast and Lite tiers exist for.

Frequently Asked Questions

What is Veo 3.1?

Veo 3.1 is Google DeepMind's video generation model. It arrived in paid preview on 15 October 2025 and reached general availability on Vertex AI on 17 November 2025, with a feature update in January 2026 that added 4K and upsampling. It generates video and audio together, and it accepts reference images to guide characters, products, and scenes.

Is Veo 3.1 still the current Veo model?

Yes — and it's the only one. Google shut down Veo 3 and Veo 2 on 30 June 2026, so Google Veo 3.1 isn't the newcomer in the family, it's the default. It has had roughly nine months of production use behind it, which is a reasonable thing to want from a model you're spending ad budget through.

Can I make ads with Veo 3.1?

That's the point of it here. Google's own customers use Veo 3.1 for TV and digital spots; on HeyOz it comes out on-brand and sized for Meta, TikTok, and more.

Is Veo 3.1 free to try?

Start on HeyOz for free — no card up front. Our pricing page has the detail on each plan.

How long can the videos be?

4, 6, or 8 seconds per generation — 8 seconds is the hard cap, and one generation returns one video. To go longer, use scene extension to continue a clip you've already made. Google describes extension as reaching well past a minute but hasn't published a firm ceiling, so plan your edit around the clips you get rather than a promised total.

What resolution does Veo 3.1 output?

720p, 1080p, and up to 4K, all at 24fps. Two caveats worth having up front: Google's own documentation isn't consistent about the 4K tier, so treat it as availability-dependent rather than guaranteed on any given run; and scene extension renders at 720p regardless, which means you can't build a long 1080p sequence out of extensions alone.

What's the difference between Veo 3.1, Fast, and Lite?

Three tiers of the same model. Standard is the full version, Fast trades some fidelity for turnaround and cost, and Lite is the cheapest for volume. The trade-off matters: Lite doesn't do reference images, scene extension, or 4K, so draft wide on Fast or Lite and finish on standard when the shot needs references or a higher tier.

What inputs does Veo 3.1 accept?

Text and images. Prompts run up to 1,024 tokens, and you can supply up to three reference images or a first and last frame. Every output carries Google's SynthID watermark, and Google holds generated files on its servers for two days — so download what you want to keep.

Every leading model, on HeyOz

Pick the right engine for the job — all in one place.

Seedance 2.0 Seedance 2.5 Google Omni Flash Seedance 1.5 Pro Wan 2.6 Kling 3 Kling O3 Kling O3 Pro GPT-Image-2 Nano Banana 2 Nano Banana Pro

Why HeyOz

Anyone can give you Veo 3.1. Only HeyOz turns it into your ad.

On-brand by default

HeyOz reads your brand and keeps every asset on-color, on-voice, on-message.

Every model, one place

Seedance, Veo, Kling, GPT Image, and more — no juggling subscriptions.

Ad-ready output

Sized and formatted for Meta and TikTok, straight out of the box.

A whole workflow

Not just a model — templates, avatars, and an agent that does it all.

Every ad your brand needs, in one place

Start Now. No agency, no brief, no blank screen.

Start creating now