The thing stopping most people from using AI video generation isn't the technology. It's the cost. The results are impressive enough. The price-per-clip makes it hard to justify at any real volume.
Google launched Veo 3.1 Lite on March 31, and the price point is the story.
$0.05 per second at 720p. $0.08 per second at 1080p.
A 30-second clip at 720p costs $1.50. At 1080p, $2.40. That's a different category of product from where AI video was six months ago.
What Veo 3.1 Lite actually is
Veo 3.1 Lite is Google's most cost-effective video model. It supports:
- Text-to-video — describe what you want, get a clip
- Image-to-video — animate a still image
- Resolutions: 720p or 1080p
- Aspect ratios: 16:9 (landscape) or 9:16 (portrait / vertical)
- Duration: 4, 6, or 8 seconds per clip — cost scales with duration
The speed is the same as Veo 3.1 Fast, the previous model in the lineup. The cost is less than half.
The pricing, broken down
| Duration | 720p | 1080p |
|---|---|---|
| 4 seconds | $0.20 | $0.32 |
| 6 seconds | $0.30 | $0.48 |
| 8 seconds | $0.40 | $0.64 |
Run 100 clips in a month and you're looking at $20–64 depending on resolution and duration. That's a realistic number for a content workflow, not an enterprise budget line.
For context: before models like this, AI video generation was priced out of reach for individual creators and small teams. The use cases existed. The economics didn't. Veo 3.1 Lite tips the balance.
Vs Veo 3 / Veo 3.1 Fast
The Veo product line now has distinct tiers:
- Veo 3.1 Fast — higher cost, same speed as Lite
- Veo 3.1 Lite — less than 50% the cost of Fast, same generation speed
- Veo 3 — the premium tier, audio support, higher quality
If you don't need audio in your clips, Veo 3.1 Lite produces comparable output at a fraction of the price. For most content use cases — background videos, product demos, social clips — the difference isn't worth paying for.
(On April 7, Google also cut prices on Veo 3.1 Fast. The whole lineup is getting more accessible.)
Where to access it
Veo 3.1 Lite is available via:
- Gemini API — for developers building programmatically
- Google AI Studio — for testing and one-off generation
Both require a paid tier account. Not available on the free tier.
What you can actually do with this
The 8-second maximum per clip sounds limiting. In practice, most effective short-form video is assembled from multiple short clips rather than one continuous take. 8 seconds is enough for a scene.
Practical use cases at this price point:
Content repurposing — take a written post or podcast excerpt and generate a visual clip to accompany it. Cost per piece: under $0.50.
Social video — portrait-mode output for Instagram Reels or TikTok. 6-second clips at $0.30 each. Viable at volume.
Product demos — animate product imagery rather than shooting video. Particularly useful if you're selling digital products with no physical asset to film.
Background visuals — ambient video for presentations, course content, or website backgrounds. Generate once, reuse everywhere.
What it can't do (yet)
- Maximum 8 seconds per clip — for longer content you're stitching
- No native audio generation (that's Veo 3) — you'll add music or voiceover separately
- Generation isn't instant — expect a wait similar to image generation at volume
- Output quality varies — as with all generative video, results are better with specific, detailed prompts than vague ones
None of these are dealbreakers. They're constraints to design your workflow around.
The gap between "AI video is a cool demo" and "AI video is part of my actual content workflow" just got a lot smaller. At $0.05 per second, the question isn't whether the economics work — they do. The question is whether you have a workflow to plug it into.
That's the more interesting problem to solve.