Grok Imagine Video 1.5 vs Veo 3.1 (Quality)

What each one accepts

Rates and clip costs

Charged from a prepaid USD balance, no taxes added. A failed generation is $0.

Grok Imagine Video 1.5 brings one photo to life as a 6-15 s clip with native sound, at 480p or 720p, billed per second. Veo 3.1 (Quality) has a fixed 8-second length, starts from a prompt or from images, renders up to 4K and is billed per clip. With a photo of a speaker in hand I choose Grok: the same 8 seconds at 720p come to roughly 1.8 times less than one Veo clip. No photo, or a need for 1080p and 4K, points to Veo.

When to pick Grok Imagine Video 1.5

• You want a talking portrait or a moving cover made from a photo you already have.
• The spoken line and its audio should come baked into the file.
• The shot from your photo needs more than 8 s, up to 15 s.
• The video is for a square 1:1 feed.

When to pick Veo 3.1

• You only have a text description.
• The ad or trailer needs 1080p or 4K.
• Several images should act as frames for a transition.
• You assemble a sequence from equal 8-second pieces.

Differences you will notice

• Starting point. Grok Imagine Video 1.5 works only from an image: exactly one, in image_url. Veo 3.1 (Quality) needs nothing but a prompt, and optionally takes up to three pictures in image_urls that serve as frames.
• Length. Grok lets you pick any whole second from 6 to 15. Veo's duration is locked, so every job is one 8-second clip at one price.
• Price. Grok bills from $0.088 per second in 480p and twice that in 720p. Veo bills from $2.53 per video for a 720p or 1080p clip, and half as much again for 4K. A minimal 6-second Grok job comes out near a fifth of a Veo clip in 480p and near half in 720p. Even 10 seconds of 720p Grok stays around 70% of one Veo clip; you need the full 15 seconds at 720p to pass it, and only just.
• Audio. Grok writes speech and sound in sync with the picture; turning generate_audio off saves nothing. The Veo form on this site has neither a sound toggle nor a sound fee. Google describes Veo 3.1 as generating audio together with video, but listen to a test clip first.
• Format and prompt. Grok: 16:9, 9:16 and 1:1, prompts up to 10,000 characters. Veo: 16:9 and 9:16, up to 4,000 characters. Anything above 720p is Veo-only.

How to decide

My test from September 28 put Grok Imagine Video 1.5 on top at an 8.5/10 average (coffee scene 8.0, tricky Polish line 9.0), with Veo 3.1 at 8.0 for both tasks. Both ran through other providers' APIs, and since the Veo form here has no audio toggle, I would not transfer the speech result one to one. To price your exact duration, open the cost calculator. Sample budgets for each model are in the Grok Imagine guide and the Veo 3.1 pricing guide; the full ranking lives in Best AI video generator 2026.

Questions

Which one handles dialogue better?

Grok did in my test: two separate speech recognition models transcribed the hard Polish sentence from its clip without a single error. You need a photo of the speaker, though.

How do I get 10 seconds out of Veo?

You cannot in one clip. Ten seconds means two 8 s clips, trimmed in editing. Grok bills 10 seconds directly at its per-second rate.

Is there a text-only option from xAI?

Yes, Grok Imagine 1.0 Video makes clips from a prompt alone. Version 1.5 always needs a starting image.

Samples from the makers

These clips and images come from the vendors, not from my runs.