Veo 3.1, Google's flagship AI video model, does not have one price. It has four, because Google sells access through four separate doors and each door meters the cost differently. There is the Gemini consumer app on a subscription, Google Flow on a credit system, the Vertex AI developer API billed per second, and per-clip access on third-party canvases like 8frame. The same 5-second clip can cost you anywhere from a fraction of a monthly subscription to $6 on the raw API, depending on which door you walk through.
This guide lays out all four routes with verified current rates, then does the campaign math so you can see which one fits your volume. If you only want the definition of the model itself, what is Veo 3 covers that. This is the money article.
The four ways to pay for Veo 3.1
1. The Gemini app (subscription)
The consumer route is a Gemini subscription. The free tier gives limited Veo access through a small monthly credit allowance. Google AI Pro, at roughly $19.99 a month, unlocks Veo 3.1 generation with around 1,000 monthly AI credits. Google AI Ultra, at roughly $249.99 a month, is the top consumer tier with a far larger credit pool and priority access to the newest model versions.
This route is the simplest and the best fit if Veo is one of several things you want from a Gemini subscription you were going to pay for anyway. It is a subscription, so the cost is fixed whether you generate one clip or a hundred, up to the credit ceiling.
2. Google Flow (credits)
Flow is Google's dedicated creator front-end for Veo, and it prices per generation in credits. A Veo 3.1 Lite generation costs around 10 credits, Fast costs around 20, and the full Quality preset costs around 100 credits per clip. Non-subscribers get a small pool of free daily credits to test with. Flow credits come bundled with the Google AI Pro and Ultra subscriptions above, so in practice Flow is where the Gemini-tier credits get spent on video.
The catch is the same as any credit system: the model preset you choose swings the cost by 10x, and the Pro tier's roughly 1,000 credits translate to only about 10 full-Quality videos a month. If you want the cinematic ceiling Veo is known for, that ceiling is expensive in credit terms.
3. Vertex AI (per-second API)
For developers building Veo into their own product, Google's Vertex AI exposes the model on a per-second billing model. The published rate for the Veo 3 Standard tier is around $0.75 per second of generated video, billed through your Google Cloud account. A standard 8-second clip therefore runs about $6.00. Organizations committing to minimum monthly spend can negotiate discounts in the 10 to 30 percent range.
The Gemini API offers cheaper variants for lower-quality tiers, with Fast presets around $0.10 per second and Lite around $0.05 per second at 720p, scaling up to $0.40 per second for the Standard preset. This route makes sense only if you're writing code against the model and want raw programmatic access. There is no canvas, no UI, and no chaining to other models. You build all of that yourself.
4. Credits on 8frame
The fourth route is credit-priced access on the 8frame canvas. Full Veo 3.1 runs 112 credits per 8-second clip, Veo 3.1 Fast runs 56, and Veo 3.1 Lite starts at 17 (a credit is about a cent at top-up pack rates). It's subscription-based, from $19/month for 1,000 credits, but the credits don't evaporate: unused subscription credits roll over while your plan is active and one-time top-up packs never expire. Generation takes about 90 seconds and outputs native 4K at 60fps with synchronized audio.
The reason this route exists alongside Google's own is that Veo rarely gets used alone. Real production chains it with other models: a reference still from Nano Banana Pro, a cheaper draft pass on Kling, an upscale on Topaz. On a multi-model canvas you run Veo next to Kling 3.0, Seedance 2.0, and 13 other models on the same surface, route each shot to the right model, and draw every clip from the same credit pool. Google's routes give you Veo and only Veo.
Pricing math by route
Here is the same 5-second clip priced four ways, using current mid-2026 rates. Ranges reflect the model presets and subscription context.
| Route | How it meters | Cost of one 5s clip | Best for |
|---|---|---|---|
| Gemini app (Pro) | Subscription + credits | ~$20/mo flat, ~10 Quality clips | Occasional creators already on Gemini |
| Google Flow | Credits per generation | ~100 credits (Quality) | Creators inside the Google ecosystem |
| Vertex AI | Per second | ~$3.75 (Standard, 5s) | Developers building on the API |
| 8frame | Credits, one pool for 16 models | 17 to 112 credits/clip by Veo tier | Multi-model production, uneven volume |
The Vertex number surprises people. At $0.75 per second, the raw API is the most expensive way to generate a finished clip unless you're operating at negotiated enterprise volume. It's priced for products that resell generation, not for teams making ads.
Worked example: a 30-second product spot
Say you're cutting a 30-second product video from six 5-second Veo clips, and you generate roughly three variants of each shot before picking winners. That's 18 generations to land 6 keepers.
On Vertex AI: 18 clips at 5 seconds and $0.75 per second is about $67.50 in compute for the spot, before any upscaling or retries. Predictable, but you're paying premium API rates for every throwaway draft.
On Google Flow (Pro tier): 18 Quality generations at ~100 credits each is 1,800 credits, which exceeds the ~1,000 monthly credits the Pro plan includes. You'd need the Ultra tier or a top-up to finish one spot at full quality in a month.
On 8frame: 18 full-Veo clips at 112 credits each is about 2,000 credits for the spot, roughly $20 at pack rates. In practice you'd draft the 12 losing variants on Veo 3.1 Lite at 17 credits and only run the 6 winners on full Veo, dropping the real cost under 900 credits, inside a single $19 Starter month. That draft-cheap, finish-expensive move is only available when the tiers live on the same canvas.
The single biggest lever on Veo spend is not the route. It's not generating expensive Veo drafts you're going to throw away. See the AI video credits explainer for the full waste-reduction sequence.
When Google's own routes are the better call
Per-clip access is not always the answer, and pretending otherwise would be dishonest. Google's routes win in specific cases.
If you're already paying for Google AI Pro or Ultra for Deep Research, Gemini, or storage, Veo access is effectively bundled and the marginal cost of a clip is zero until you hit the credit ceiling. If you're a developer building Veo into a product and need programmatic control, deterministic latency, and Google Cloud billing integration, Vertex AI is the correct route and no canvas replaces it. And if you generate a very high, steady volume of Veo specifically, the Ultra tier's large credit pool can beat per-clip rates at the top of the volume curve.
The canvas route wins on the things those cases don't cover: uneven volume, multi-model chaining, credits that roll over instead of expiring, and one plan from $19/month instead of a per-tool stack. For a fuller decision walkthrough, how to use Veo 3.1 covers every access option step by step.
FAQ
How much does Veo 3.1 cost per video?
It depends entirely on the route. On the Vertex AI API it's about $0.75 per second, so roughly $3.75 for a 5-second clip. On the 8frame canvas it's 112 credits per 8-second clip (about $1.12 at pack rates), with Veo 3.1 Fast at 56 and Lite from 17. Inside a Gemini subscription it's bundled into a monthly fee of roughly $20 (Pro) or $250 (Ultra) with credit limits.
Is Veo 3.1 available for free?
Partially. The free Gemini tier and Google Flow both give a small pool of free credits or daily allowances that let you generate a handful of clips to test the model. For any real production volume you'll move to a paid route quickly, because the free allowances are sized for evaluation, not delivery.
What is the cheapest way to use Veo 3.1?
For low, uneven volume, a pooled-credit canvas like 8frame is usually cheapest because credits roll over and you can draft on Veo 3.1 Lite from 17 credits before finishing on the full model. For high, steady volume on Veo specifically, a Gemini Ultra subscription's large credit pool can come out ahead. Run your own monthly clip count through both before committing.
Is the Vertex AI API price really $6 for one clip?
Yes, for a standard 8-second clip at the Veo 3 Standard rate of about $0.75 per second, before any enterprise discount. The API is priced for products that build on top of Veo, not for end users making individual videos, which is why credit-priced canvas access exists at a fraction of the cost.
Run Veo 3.1 alongside 15 other models from one credit pool, from $19/month. Open the 8frame canvas and generate your first Veo clip today.