Video Generation API: Models, Methods, and Billing

A video generation call on Aigenway always has the same four fields — provider, model, method, params — whether it's Veo 3 rendering with native audio or Seedance churning out a batch of cheap drafts. This page is the map: which providers are live, what text_to_video and image_to_video actually take, how the async task and webhook flow works, and how prepaid USD billing prices a clip, with links out to the model-specific guides.

The providers and models behind the video endpoint

Three providers carry video generation today: veo (Google's Veo 3, native audio), kling (motion control and character consistency), and seedance (speed and cost at volume). Each exposes its own model codes and some charge per-second while params like duration or aspect_ratio vary slightly by model, so don't hardcode a code or price from a blog post — list what's live on your account right now:

curl "https://api.aigenway.com/v1/methods?provider=veo" \
  -H "Authorization: Bearer sk_live_..."

curl "https://api.aigenway.com/v1/methods?provider=kling" \
  -H "Authorization: Bearer sk_live_..."

curl "https://api.aigenway.com/v1/methods?provider=seedance" \
  -H "Authorization: Bearer sk_live_..."

Each response lists the current model codes, the method values each one supports, and the per-unit price for your account. For a head-to-head of what each is actually good at, see Veo 3 vs Kling vs Seedance and the cheapest AI video API.

text_to_video: a clip from a prompt

text_to_video takes a prompt plus model-specific extras like duration and aspect_ratio, and renders a clip from nothing:

curl -X POST https://api.aigenway.com/v1/generations \
  -H "Authorization: Bearer sk_live_..." \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "kling",
    "model": "kling-v2-master",
    "method": "text_to_video",
    "params": {
      "prompt": "a lighthouse beam sweeping over a stormy coastline at night",
      "duration": 5,
      "aspect_ratio": "16:9"
    }
  }'

Swap provider/model for veo/veo-3 or seedance/seedance-1-pro (confirmed via /v1/methods) and the same request shape, headers, and Idempotency-Key carry over unchanged — the request envelope doesn't care which model renders it.

image_to_video and the async task you get back

image_to_video takes an existing image plus a motion-focused prompt and animates it — swap method to image_to_video, add input_image, and drop the scene description for something about the motion itself:

curl -X POST https://api.aigenway.com/v1/generations \
  -H "Authorization: Bearer sk_live_..." \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "seedance",
    "model": "seedance-1-pro",
    "method": "image_to_video",
    "params": {
      "prompt": "camera pulls back slowly, product rotates gently",
      "input_image": "https://example.com/product-shot.png",
      "duration": 5
    }
  }'

Neither method returns a finished clip inline — video renders take longer than an image call, so POST /v1/generations always hands back a task that moves through pending → processing → succeeded or failed:

{
  "task_id": "cmomwuy0m000qbr0371src20a",
  "status": "pending",
  "provider": "seedance",
  "model": "seedance-1-pro",
  "method": "image_to_video",
  "created_at": "2026-10-16T09:12:03.000Z"
}

Poll GET /v1/tasks/{task_id} or set a callback_url on the submit and let a signed webhook deliver the result when it's ready instead of holding a connection open for a multi-minute render — either path ends with the same result object once status is succeeded. result.url is a signed link that expires; download it rather than treating it as permanent storage. Full state machine on the task lifecycle page, the full webhook payload and signature on async tasks and webhooks, and the trade-off between the two on poll vs. webhook generation.

How billing actually prices a video call

Video models typically bill per_second in nano-dollars (10⁻⁸ USD — divide by 100,000,000 for a USD amount), the same ledger unit used across every provider on the gateway, not a per-vendor credit pack. duration multiplies that per-second rate directly — a 10-second clip costs roughly 2× a 5-second clip on the same model, not a flat per-render fee. Price a request before you spend on it with POST /v1/estimate, which takes the identical body as /v1/generations and returns pricing.final_price plus balance.enough_balance without charging anything or starting a render:

curl -X POST https://api.aigenway.com/v1/estimate \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "veo",
    "model": "veo-3",
    "method": "text_to_video",
    "params": { "prompt": "a drone shot over a canyon at sunrise", "duration": 8 }
  }'

The per-unit rate itself lives on /v1/methods for your account, not in this post — see estimate generation cost before you spend for the full response shape, and AI generation API pricing compared for how video pricing stacks up against image and text.

Why route video generation through one gateway

A provider-direct integration for one video model works fine until the brief needs native audio, motion control, or a cheaper draft tier the first model doesn't offer. One gateway account replaces that churn with:

• One key for Veo, Kling, and Seedance. They share the same Authorization: Bearer sk_live_... header and request shape — no separate GCP project or vendor account per model.
• One contract for text-to-video and image-to-video. method is the only thing that changes between rendering from a prompt and animating a still; the envelope, task states, and webhook format don't.
• One async flow regardless of render time. The same pending → processing → succeeded task and the same webhook signature cover a 5-second Seedance draft and an 8-second Veo 3 render with audio.
• One prepaid USD balance. Top up in dollars, spend in nano-dollars, check what's left with GET /v1/balance — no per-provider credits ledger to reconcile before forecasting a video budget.

Start from /v1/methods for the live model codes, then work through the model-specific guides — Kling, Seedance, Veo 3 — for the params each one actually uses.