Veo 3 pricing follows two paths: the subscription (Google AI Pro / Ultra, with Flow credits) and the API, billed per clip. With the API, on Kunavo, Veo 3.1 Lite costs $0.18+, Veo 3.1 Fast costs $0.36+, and Veo 3.1 Quality costs $1.60+ per 720p clip — with no monthly plan and no Google Cloud project. The cheaper path is determined by one division: subscription price ÷ price per clip = break-even point.
When you search for “how much does Veo 3 cost,” you find almost exclusively subscription comparisons. That half of the answer is covered; the other half — what each clip costs when paid for individually through the API — is missing. This page fills that gap with the numbers.
Two ways to pay for Veo 3
| Route | Billing | Limit |
|---|---|---|
| Subscription (Google AI Pro / Ultra, Flow) | Monthly plan + credits | The month’s credits |
| API (pay-as-you-go) | Per clip | No plan, no cap |
Subscription prices and the number of credits change often: we do not reproduce them here; check them on Google’s official page. For comparison, the division below is enough.
Veo 3 API pricing — per clip, by tier
The prices below are read from the model catalog: if the price changes, the table changes with it.
| Tier | Model ID | Per clip | Use |
|---|---|---|---|
| Veo 3.1 Lite | veo-3-lite | $0.18+ | Drafts, prompt tuning, volume testing |
| Veo 3.1 Fast | veo-3 | $0.36+ | Everyday renders |
| Veo 3.1 Quality | veo-3-quality | $1.60+ | Final version, key shots |
The “+” means “starting at”: it is the minimum price for a 720p clip, and a higher resolution may cost more. The IDs remain those of the veo-3 family; the current generation shown in the table runs behind the scenes. Billing is in dollars, clip by clip, against a prepaid balance.
Break-even point — subscription price ÷ price per clip
Subscription or API: the number of clips per month decides. Convert the subscription price to dollars and divide it by the price of the tier you use: you get the break-even point. Below it, the API costs less; above it, month after month, the subscription wins. As a reference, here is how many clips $10 of subscription is worth:
| Tier | Per clip | $10 of subscription ≈ |
|---|---|---|
| Veo 3.1 Lite | $0.18+ | about 55 clips |
| Veo 3.1 Fast | $0.36+ | about 27 clips |
| Veo 3.1 Quality | $1.60+ | about 6 clips |
Multiply by the number of times your subscription contains $10: you have the break-even point for your plan. One more figure is worth checking: if the subscription credits cover fewer clips than that break-even point, the API is cheaper in every case. Conversely, if you generate in Flow almost every day and use up your credits each month, stay with the subscription — especially if you also use the rest of the plan.
The strongest lever — mixing tiers
The biggest reduction in the bill comes from separating testing from the final version. Refine the prompt in Veo 3.1 Lite as many times as needed, then run only the approved prompt in Veo 3.1 Quality. With 10 drafts and 1 final version, doing everything in Veo 3.1 Quality costs $17.60; moving the drafts to Veo 3.1 Lite, $3.40. In the code, only one word changes: the model name.
The call — just watch out for asynchronous processing
Unlike text, video is an asynchronous task. POST /v1/videos first returns the task ID; you then query GET /v1/videos/{id} until the status is completed or failed.
# Le minimum pour générer un clip et attendre le résultat.
# Ni abonnement ni projet Google Cloud : une seule clé.
import os, time, requests
KEY = os.environ["KUNAVO_API_KEY"]
H = {"Authorization": f"Bearer {KEY}"}
# 1. Envoi — renvoie tout de suite l'identifiant de la tâche (statut "queued").
job = requests.post(
"https://api.kunavo.com/v1/videos",
headers=H,
json={
"model": "veo-3-lite", # brouillons en Lite, version finale en "veo-3-quality"
"prompt": "une ruelle de Montmartre sous la pluie, reflets des réverbères, la caméra avance lentement",
"aspect_ratio": "16:9",
},
timeout=60,
).json()
# 2. La génération vidéo est asynchrone : on interroge la tâche jusqu'à la fin.
while True:
r = requests.get(
f"https://api.kunavo.com/v1/videos/{job['id']}", headers=H, timeout=30
).json()
if r["status"] in ("completed", "failed"):
break
time.sleep(5)
print(r["output"]["url"] if r["status"] == "completed" else r["error"])Failed requests are not billed. Regenerating because the result is unsatisfactory, however, is a successful generation and is billed each time. Response format, image-to-video, and webhooks are all covered in the video API documentation (in English).
For payment, top-ups are processed through Stripe: bank cards (CB, Visa, Mastercard, American Express), Apple Pay, Google Pay, or Link, plus Bancontact for payments from Belgium. SEPA Direct Debit is not offered; a debit card works normally. Prepaid balance from a $10 top-up, with no subscription, and the balance never expires.
In all honesty — when this approach is not right
Through the gateway, you start without a Google billing account or Cloud project, but with shared capacity, not a quota dedicated to your project. If your volume requires a guaranteed quota or a contractual SLA, go directly through Google: this page does not claim to replace that.
The complete guide, in English, is in Veo 3 API, and the comparison with the other video models is in the text-to-video API guide.
Frequently asked questions
How much does Veo 3 cost?
Veo 3 pricing follows two models. The subscription (Google AI Pro / Ultra, with Flow credits) is a monthly plan, and generation remains limited to the month’s credits. With the API, you pay per clip: on Kunavo, Veo 3.1 Lite costs $0.18+, Veo 3.1 Fast costs $0.36+, and Veo 3.1 Quality costs $1.60+ per clip (minimum price, in 720p). No monthly plan or usage minimum.
What is the price of the Veo 3 API per video?
On Kunavo, the Veo 3 API is billed per clip, by tier: $0.18+ for Veo 3.1 Lite, $0.36+ for Veo 3.1 Fast, and $1.60+ for Veo 3.1 Quality. The “+” means “starting at”: it is the price of a 720p clip, and a higher resolution may cost more. Failed requests are not charged.
Subscription or API: which is cheaper for Veo 3?
It depends entirely on the number of clips you generate per month. The calculation is one division: subscription price (in dollars) ÷ price per clip = break-even number of clips. In months when you generate fewer, the API costs less; if you generate more each month consistently, the subscription credits are better value. And if the subscription credits cover fewer clips than that break-even point, the API is cheaper in every case.
Can you use the Veo 3 API without a Google Cloud project?
Yes, through an OpenAI-compatible gateway: no Google billing account, no project to create, and one key. The trade-off is shared capacity rather than a quota dedicated to your project. If you need a guaranteed quota or a contractual SLA, go directly through Google.
What is the difference between Veo 3 Lite, Fast, and Quality?
Speed, image quality, and price per clip. Lite is the cheapest tier, for drafts and high-volume testing; Fast is suitable for everyday renders; Quality is for the final version. To spend the least, refine the prompt in Lite and run only the approved prompt in Quality: the difference between $0.18+ and $1.60+ corresponds exactly to the cost of your tests.
Is API video generation instant?
No. Video is an asynchronous task: POST /v1/videos first returns the task ID, then you query GET /v1/videos/{id} until the status changes to completed or failed. Unlike text models, there is no single response: the code separates submission and retrieval.
Are failed requests billed?
On Kunavo, failed requests are not billed. Regenerating because you do not like the result, however, counts as a successful generation and is billed each time. That is why we recommend the cheapest tier until the prompt is finalized.