Polaris
AI modelsVeo 3.1 for Ads
AI model

Veo 3.1 for Ads

Cinematic 8-second ads with native spoken dialogue — and a Fast tier built for hook testing

Veo 3.1 is the cinematic option in Polaris, and the only video model that speaks native dialogue with no sound add-on. It renders 8-second clips in 9:16 or 16:9 across two tiers: Quality (1080p = 200, 4K = 320 credits) and Fast (720p = 41, 1080p = 49, 4K = 120 credits).

What Veo 3.1 is best at

Two things. First, the look: Veo leans filmic — lighting, depth, and framing that read closer to a produced spot than a phone clip. When you want an AI ad that feels premium rather than casual, Veo is the pick. Second, the dialogue: because speech is generated natively inside the render, a line like a customer reacting to your product comes out synced without any add-on or extra credits.

Quality vs Fast: how to split the work

Tier720p1080p4K
Quality (8s)200320
Fast (8s)4149120

The workflow that makes sense for most stores: draft on Fast, finish on Quality. At 41 credits per 720p clip, Fast is the cheapest way in the Polaris lineup to test many hooks — run a parallel batch of variations, see which opening lands, then re-render the winner on Quality for the final asset. See pricing for how far a monthly credit pool stretches across both tiers.

Trade-offs to know

Veo is fixed at 8 seconds — no 5s quick cut, no 10s slow build like Kling offers. There is no 1:1 ratio, so square placements are off the table. And the Quality tier is the most expensive render per clip in Polaris, which is fine for finals but wasteful for exploration — that is exactly what Fast is for. As always, the cinematic look is a hypothesis, not a guarantee; test it against a plainer UGC style before scaling spend.

How to use it

  1. With the Polaris MCP connected, ask Claude for the ad in plain chat — product photo attached or script pasted.
  2. On the tappable settings cards, pick Veo 3.1, choose Quality or Fast, then resolution and ratio.
  3. The live panel shows progress and the finished clip in 20-180 seconds; failed renders auto-refund.

The full setup is covered in how to generate ads inside Claude with MCP.

Frequently asked questions

What is the difference between Veo 3.1 Quality and Fast?
Both render 8-second clips in 9:16 or 16:9. Quality costs 150/178/200 credits at 720p/1080p/4K and is the finishing tier; Fast costs 41/49/120 credits and is the cheap way to test many hook variations before committing.
Does Veo 3.1 need the AI sound add-on?
No. Veo 3.1 speaks native dialogue inside the render itself, so spoken lines come out synced with no add-on and no extra credits.
Can Veo 3.1 render square video?
No — Veo 3.1 supports 9:16 and 16:9 only. If you need 1:1 for square placements, Kling 3.0 is the only Polaris model that offers it.
When should I pick Veo 3.1 over Kling 3.0?
Pick Veo when you want a cinematic, produced look or native spoken dialogue, or when you want to test lots of hooks cheaply on the Fast tier. Pick Kling for all-round natural UGC motion, 5s/10s durations, or 1:1.

Generate winning ads inside Claude

Connect Polaris to Claude and generate UGC-style video and image ads right in the chat. First batch of 12 free.

AI models

GPT Image 2 for Ad ImagesKling 3.0 for AdsNano Banana Pro for Ad ImagesOmni Talking-Actor Model

Related reading

GuideGenerate ecommerce ads inside Claude (MCP guide)GuideImage to video: turn a product photo into a video ad