Answers

How much does one AI-generated video actually cost?

By BananaBanana TeamPublished Last checked

Short answer

Between $0.10 and $4.40 per clip, depending on model, resolution, length and whether the audio track is generated. A 4-second silent 720p clip on Veo 3.1 Lite is $0.10; an 8-second 4K clip with audio on Veo 3.1 is $4.40. Gemini Omni Flash bills per second — $0.10 a second at 720p, so $0.30 to $1.00 per clip. You pay per generation, not per month.

Cheapest clip
$0.10 — Veo 3.1 Lite, 4 s, 720p, silent
Most expensive clip
$4.40 — Veo 3.1, 8 s, 4K, with audio
Veo 3.1 Fast, 8 s, 1080p, audio
$1.00
Gemini Omni Flash
$0.10 per second at 720p, audio always included
Billing model
Per generation, prepaid balance, no subscription
Failed generation
Refunded in full, automatically

Facts on this page were checked against the live platform on .

01In detail

What drives the price of a clip?

Four things, in order of impact: the model, the resolution, the duration, and whether audio is generated. Everything else — aspect ratio, first frame, reference images, negative prompt — is free.

Audio is the surprise for most people. On Veo 3.1 it roughly doubles the price: the same 8-second 720p clip is $1.40 silent and $3.00 with a native soundtrack. On Gemini Omni Flash there is no choice to make — audio is always generated and already in the per-second rate.

Resolution is not linear either. Going 720p → 1080p is free on Veo 3.1 and Veo 3.1 Fast (same price row), while 4K costs roughly twice as much as 1080p. On Veo 3.1 Lite there is no 4K at all, and 1080p carries a modest premium over 720p.

Which model should I pay for?

Veo 3.1 Lite ($0.10–$0.36) is the draft tier: use it to find out whether the shot works at all. Veo 3.1 Fast ($0.35–$2.60) is the everyday choice — it keeps most of the motion quality at a third of the flagship price. Veo 3.1 ($0.70–$4.40) is worth it when the clip is the deliverable: complex camera moves, dialogue, or a 4K master.

Gemini Omni Flash ($0.30–$1.00) is a different tool rather than a cheaper Veo. It always makes 3–10 second clips with synchronised audio and accepts conversational edits — you ask for a change to a clip you already made instead of re-rolling the prompt. For talking-head and short-form social shots it usually beats a silent Veo clip plus a separate voice-over.

What does a finished piece cost, not just one clip?

A useful rule from real projects: budget three to five generations per usable shot. Prompt drift, a hand in the wrong place, a filter rejection — none of it is unusual, and filter rejections at least come back as a refund.

So a 30-second social spot cut from six shots realistically means 18–30 generations. On Veo 3.1 Fast at 8 seconds, 1080p, with audio, that is $18–$30 of compute. The same piece drafted on Veo 3.1 Lite first and finished on Fast typically lands under $12.

There is no subscription, minimum, or expiring credit on top of that. Balance is topped up in cryptocurrency, and top-ups above $50 carry a +5% bonus, above $100 a +10% one.

02Go deeper

Where this is documented

03Guides

Longer reads on the same thing

04Also asked

More on this question

Is there a monthly plan?

No. Balance is prepaid and spent per generation; an idle month costs nothing and unspent balance does not expire.

Do I pay when a generation fails?

No. Upstream failures, capacity errors and content-filter rejections are refunded automatically. A partially filtered batch is refunded proportionally.

Is extending a clip cheaper than generating a new one?

It is priced as its own generation. A Veo extension adds 7 seconds and is billed as a 7-second, single-variant clip on the same model.

How does this compare to a credits-based platform?

The difference is that a dollar here is a dollar of generation. Credit systems bundle a monthly fee, round each generation up to whole credits, and usually expire the unused ones.

05Related

Questions next door

Try it on your own prompt

New accounts start with $0.20 of balance — no card, nothing expires.