Veo 3 API pricing per 8-second video: every provider
The 8-second clip is the natural unit for comparing Veo 3 API costs — it is the standard generation length across providers. Here is what each one charges for it, as of 2026:
| Provider | Price per 8s video | Price per second | Billing model |
|---|---|---|---|
| Google Vertex AI | $2.88 | $0.36 | GCP metered billing |
| fal.ai | $3.20 | $0.40 | Per-call, prepaid balance |
| Replicate | $6.00 | $0.75 | Per-second of output |
| Veo3Gen — Fast | $0.55–$0.83 | ≈ $0.07–$0.10 | 10 credits per video |
| Veo3Gen — Quality | $1.43–$2.16 | ≈ $0.18–$0.27 | 26 credits per video |
| Veo3Gen — Veo 3.1 Lite | from $0.17 | ≈ $0.02–$0.05 | 3–5 credits per video |
Every row above serves the same underlying Google Veo model family, so the spread — from $0.17 to $6.00 for a functionally identical video — comes entirely from how each provider packages access. Vertex AI bills raw compute through Google Cloud; fal.ai and Replicate resell it per call with a margin; Veo3Gen buys capacity in volume and resells it as flat-priced credits.
Vertex AI Veo 3 pricing in 2026, explained
Google's first-party route is Vertex AI, which meters Veo 3 at $0.36 per second of generated video — $2.88 per 8-second clip, $1.44 for 4 seconds. The rate is the same whether the generation is usable or not: every retry, every prompt experiment, every discarded take bills at full price. You also need a Google Cloud project with billing enabled, a service account, and quota approval before your first request.
For teams already deep in GCP with committed-use discounts, that overhead may be acceptable. If you want the full breakdown of Google's official rates — including how Vertex quotas and billing tiers work — see our dedicated guide to official Google Veo API pricing. This page focuses on the cross-provider picture.
fal.ai and Replicate: convenience at a markup
fal.ai and Replicate remove the GCP setup pain — simple REST endpoints, no cloud console — but you pay for the convenience. fal.ai lists Veo 3 at $3.20 per 8-second video($0.40/s), about 11% above Vertex AI's own rate. Replicate charges $0.75 per second, which works out to $6.00 per 8-second clip — more than double Google's first-party price for the identical model.
That pattern is worth internalizing: with resold models, developer experience and price usually move in opposite directions. Veo3Gen's bet is that they don't have to — a simple API key and a documented REST endpoint, priced below even Google's own metered rate through volume credit pricing.
Veo3Gen API pricing: how credits convert to dollars
Veo3Gen bills in credits instead of metered seconds, which makes costs predictable per call. Each generation deducts a fixed amount depending on the model:
| Model | Credits per 8s video | Effective cost | Best for |
|---|---|---|---|
| Veo 3.1 Lite (720p–1080p + audio) | 3–5 | $0.17–$0.42 | Drafts, high-volume products |
| Veo 3 Fast (720p/1080p + audio) | 10 | $0.55–$0.83 | Production default |
| Veo 3 Quality (720p/1080p + audio) | 26 | $1.43–$2.16 | Hero shots, client work |
| 4K on Veo 3.1 (Fast / Quality) | 22 / 38 | $1.21–$3.15 | Premium deliverables |
Credits come from one-time packs — Basic $9.99 (120 credits), Hero $37.50 (450), Studio $79.99 (1,000) — or monthly plans with better rates: Basic $9.99/180, Pro $37.50/600, Studio $79.99/1,200. That puts the effective cost per credit between $0.055 and $0.083, which is where the per-video ranges above come from. Credits are valid at least 30 days from purchase (see Terms), and full plan details are on the pricing page.
Run the numbers on your own workload — API keys take under a minute.
Get API AccessWhat real projects cost: 10, 100, and 1,000 videos
Per-video pricing only becomes real at project scale. Using each provider's 8-second rate and Veo3Gen's Fast model:
| Volume | Vertex AI | fal.ai | Replicate | Veo3Gen (Fast) |
|---|---|---|---|---|
| 10 videos | $28.80 | $32.00 | $60.00 | ≈ $8.30 (Basic pack) |
| 100 videos | $288.00 | $320.00 | $600.00 | $79.99 (Studio pack, 1,000 credits) |
| 1,000 videos/mo | $2,880 | $3,200 | $6,000 | ≈ $667/mo (Studio plans) |
The 100-video row is the cleanest illustration: a single $79.99 Studio pack covers exactly 100 Fast generations that would cost $288 on Vertex AI or $600 on Replicate — a 72–87% saving before you account for retries. Switch draft iterations to Veo 3.1 Lite at 3 credits each and the gap widens further, since experimentation stops costing production prices.
How to choose a Veo 3 API provider
Since the model output is identical everywhere, the decision reduces to billing fit. Vertex AI makes sense if your stack already lives in Google Cloud and procurement requires a first-party vendor. fal.ai or Replicate fit teams that want one aggregator account across many models and accept the markup. Veo3Gen is built for the common case: you want Veo 3 in your product at the lowest per-video cost, with flat credit pricing you can put in a spreadsheet.
If you are new to the API, start with what the Veo 3 API is and the getting-started guide; for a deeper head-to-head with Google's platform, see Veo 3 API vs Google Vertex AI and why cheap Veo 3 API access works. When you are ready, grab an API key — no credit card required to sign in and look around.
Frequently asked questions
How much does the Veo 3 API cost on Vertex AI?
What is the cheapest Veo 3 API in 2026?
Do all Veo 3 API providers run the same model?
Is there a free way to test the Veo 3 API?
How does Veo3Gen API billing work?
Does Veo 3 API pricing change with video length or resolution?
Stop paying $2.88 per video for a $0.55 generation
Vertex-grade Veo 3 output at credit-pack prices — from $9.99, valid 30+ days, no GCP project required.
No credit card required — sign in with Google and start in seconds.