Veo 3.1 - AI Model API and Pricing

Jobs that justify the rate

Veo 3.1 generates audio with the video, dialogue included, and follows detailed prompts closely. That pays off on a narrow set of jobs. Think of a spokesperson line for an ad or a campaign hero shot that opens a launch film. A short scene where two people trade a sentence fits too. For scenic B-roll with no speech, you pay for abilities nobody will hear.

A dialogue example: "Close-up of a barista in her thirties looking into the lens. She says 'We roast every Monday, so it never sits on the shelf'. Café noise behind her, soft window light." Keep spoken lines short so they fit the clip.

Which tool and preset to open

In AI Video Generator, the Talking to camera preset is the natural start for spoken lines. Product turn or Slow push in suit hero shots of a single object. Cinematic build and Drone reveal work for openers. Text to Video leans scenic, with presets such as Coffee pour, City timelapse, Fashion walk and Ocean sunrise. Neither tool takes an upload, so the shot starts from your words.

A hero-shot prompt: "Slow push in on a matte black headphone case opening on a marble table. Morning light, a soft click as the lid lifts."

Settings: length, resolution, frame

You can render 4, 6 or 8 seconds at 720p, or 8 seconds at 1080p, in 16:9 or 9:16. Use 9:16 for Reels and Shorts, 16:9 for web and TV. Choose the length from the dialogue. A short line fits in 4 seconds, while a full sentence plus a reaction needs 8. Because billing is per second, render the shortest length the line allows.

A 4-second clip costs 380 credits ($4.04 at the best pack). A 6-second clip costs 570 ($6.07), and an 8-second clip costs 760 ($8.09).

Cheaper routes to the same shot

Sorted from cheapest, Veo 3.1 sits 30th of 31 models in its group. Drafting on it gets expensive fast. Veo 3.1 Fast at 36 per second keeps Veo behavior with less detail. Lock wording, framing and timing there. Then rerun only the approved take on the full model. Kling 3.0 Pro at 40 is the better finisher when a project already uses the Kling look. Wan 2.6 at 18 suits longer clips where fidelity comes second.

The real limit is cost. A wrong guess on an 8-second master burns 760 credits, so keep exploration on a cheaper sibling.

Where to actually run it

Veo 3.1 API

curl -X POST https://ugesi.com/api/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "veo-3-1",
    "prompt": "a product shot on white marble",
    "seconds": 4
  }'

<?php
$ch = curl_init('https://ugesi.com/api/generate');
curl_setopt_array($ch, [
    CURLOPT_POST => true,
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_HTTPHEADER => [
        'Authorization: Bearer YOUR_API_KEY',
        'Content-Type: application/json',
    ],
    CURLOPT_POSTFIELDS => json_encode([
        'model' => 'veo-3-1',
        'prompt' => 'a product shot on white marble',
        'seconds' => 4,
    ]),
]);
$response = json_decode(curl_exec($ch), true);
curl_close($ch);
print_r($response);

type GenerateRequest = {
  model: string;
  prompt: string;
  seconds?: number;
};
type GenerateResponse = {
  job_id: number;
  job_token: string;
  credits_held: number;
  status: "queued" | "succeeded";
  output_text?: string;
};

const body: GenerateRequest = {
  model: "veo-3-1",
  prompt: "a product shot on white marble",
  seconds: 4,
};

const resp = await fetch("https://ugesi.com/api/generate", {
  method: "POST",
  headers: {
    "Authorization": `Bearer ${process.env.UGESI_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify(body),
});
if (!resp.ok) throw new Error(`HTTP ${resp.status}`);
const job = (await resp.json()) as GenerateResponse;
console.log(job.job_id, job.status);

import requests

resp = requests.post(
    "https://ugesi.com/api/generate",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "model": "veo-3-1",
        "prompt": "a product shot on white marble",
        "seconds": 4,
    },
    timeout=60,
)
resp.raise_for_status()
print(resp.json())

{
  "model": "veo-3-1",
  "prompt": "a product shot on white marble",
  "seconds": 4
}

The API comes with every credit pack, from Lite to Enterprise, and uses the same balance as the website. API documentation and access →

License and terms of use

Outputs made with paid credits can be used commercially. The maker's own rules for Veo 3.1 also apply (Google Generative AI Prohibited Use Policy). The full rules are in our Terms of service.

People also look for