Yapper Developers
Models & Inputs

Video generation

Video models with per-length pricing and input fields.

POST /api/v1/processes with "type": "video-generation". Pick any model — the API does not plan-gate models. GET /api/v1/models returns the same capabilities at runtime.

Input fields

FieldTypeRequiredDescription
promptstringyesWhat to generate.
aspectRatiostringnoOne of the model's aspect ratios, e.g. "16:9".
endingFrameImageUrlstringnoImage URL used as the last frame. External URLs are auto-imported as assets first.
referenceImages{docId, url}[]noYapper image assets. Use the id + url from /assets, a process output, or /assets/import — passing docId preserves asset lineage.
referenceVideos{docId, url}[]noYapper video assets (same docId + url shape).
resolutionnumbernoOutput resolution height, e.g. 1080.
startingFrameImageUrlstringnoImage URL used as the first frame. External URLs are auto-imported as assets first.
videoLengthnumbernoDuration in seconds.

Referencing media: referenceImageUrls takes plain https URLs (external images). The reference* asset fields take docId + url pairs pointing at Yapper assets — use the id and url returned by GET /assets, a completed process output, or POST /assets/import. Prefer the asset form when the media lives in Yapper: the docId preserves asset lineage.

Example

Add "dryRun": true to get the exact credit cost without starting, and always send an Idempotency-Key header on real starts.

POST /api/v1/processes
{
  "type": "video-generation",
  "model": "grok-imagine-v1.5",
  "input": {
    "prompt": "A slow dolly shot over a foggy pine forest at sunrise",
    "videoLength": 1,
    "aspectRatio": "16:9"
  }
}

Models at a glance

Credits are per generated output; ranges span the supported resolutions/lengths (exact prices in each model's section).

ModelCompanyCreditsLengths (s)Max ref imagesFramesAudio
seedance-2.0bytedance200–7604–159yes
grok-imaginexai58–1354, 6, 8, 10, 12, 14, 15startyes
grok-imagine-v1.5xai20–2601–15startyes
happy-horsealibaba90–4503–155yes
seedance-2.0-fastbytedance180–6804–159yes
wan-2.7alibaba30–1502, 4, 6, 8, 105yes
pixverse-v6pixverse20–3701, 2, 4, 6, 8, 10, 12, 14, 15startyes
kling-3.0kuaishou72–2704, 6, 8, 10, 12, 14, 153start + endyes
kling-3.0-prokuaishou96–3604, 6, 8, 10, 12, 14, 153start + endyes
gemini-omni-flashgoogle80–2503–10yes
veo3-qualitygoogle21683start + endyes
veo3-fastgoogle8083start + endyes
sora-2openai48–2404, 8, 12, 16, 20startyes
sora-2-proopenai96–4804, 8, 12, 16, 20startyes

Seedance 2.0 (seedance-2.0)

ByteDance's video model with multi-modal input, lip sync, and multi-shot narrative.

Credits4s: 200 · 5s: 250 · 6s: 300 · 7s: 350 · 8s: 400 · 9s: 450 · 10s: 500 · 11s: 550 · 12s: 610 · 13s: 660 · 14s: 710 · 15s: 760 (at 1080p)
Aspect ratios16:9, 4:3, 1:1, 3:4, 9:16, 21:9
Lengths (s)4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Resolutions480, 720, 1080, 2160
Reference imagesup to 9
Reference videosup to 3 (15s total)
Reference audioup to 3 (15s total)
Audio outputyes
Max prompt length20000

Grok Imagine (grok-imagine)

Grok's best and latest video model with sound included.

Credits4s: 58 · 6s: 72 · 8s: 86 · 10s: 100 · 12s: 114 · 14s: 128 · 15s: 135 (at 720p)
Aspect ratios16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 1:1 — "auto" also accepted with startingFrameImageUrl
Lengths (s)4, 6, 8, 10, 12, 14, 15
Resolutions720
Start/end framesstart
Audio outputyes
Max prompt length4080

Grok Imagine 1.5 (grok-imagine-v1.5)

xAI's Grok Imagine 1.5 image-to-video model with audio included.

Credits1s: 20 · 2s: 40 · 3s: 50 · 4s: 70 · 5s: 90 · 6s: 110 · 7s: 120 · 8s: 140 · 9s: 160 · 10s: 170 · 11s: 190 · 12s: 210 · 13s: 230 · 14s: 240 · 15s: 260 (at 480p)
Aspect ratios16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 1:1 — "auto" also accepted with startingFrameImageUrl
Lengths (s)1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Resolutions480, 720
Start/end framesstart
Audio outputyes
Max prompt length4096

Happy Horse (happy-horse)

Alibaba's reference-to-video model with native audio and 720p output.

Credits3s: 90 · 4s: 120 · 5s: 150 · 6s: 180 · 7s: 210 · 8s: 240 · 9s: 270 · 10s: 300 · 11s: 330 · 12s: 360 · 13s: 390 · 14s: 420 · 15s: 450 (at 720p)
Aspect ratios16:9, 9:16, 1:1, 4:3, 3:4
Lengths (s)3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Resolutions720
Reference imagesup to 5
Audio outputyes
Max prompt length5000

Seedance 2.0 Fast (seedance-2.0-fast)

Fast variant of ByteDance's video model with multi-modal input, lip sync, and multi-shot narrative.

Credits4s: 180 · 5s: 230 · 6s: 270 · 7s: 320 · 8s: 360 · 9s: 410 · 10s: 450 · 11s: 500 · 12s: 540 · 13s: 590 · 14s: 640 · 15s: 680 (at 1080p)
Aspect ratios16:9, 4:3, 1:1, 3:4, 9:16, 21:9
Lengths (s)4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Resolutions480, 720, 1080, 2160
Reference imagesup to 9
Reference videosup to 3 (15s total)
Reference audioup to 3 (15s total)
Audio outputyes
Max prompt length20000

WAN 2.7 (wan-2.7)

Wan's latest reference-driven video model with image and video guidance for better consistency.

Credits2s: 30 · 4s: 60 · 6s: 90 · 8s: 120 · 10s: 150 (at 1080p)
Aspect ratios16:9, 9:16, 1:1, 4:3, 3:4
Lengths (s)2, 4, 6, 8, 10
Resolutions720, 1080
Reference imagesup to 5
Reference videosup to 1
Audio outputyes
Max prompt length5000

PixVerse V6 (pixverse-v6)

PixVerse's latest Fal-hosted video model with optional audio and first-frame image support.

Credits1s: 20 · 2s: 50 · 4s: 100 · 6s: 150 · 8s: 200 · 10s: 250 · 12s: 300 · 14s: 350 · 15s: 370 (at 1080p)
Aspect ratios16:9, 4:3, 1:1, 3:4, 9:16, 2:3, 3:2, 21:9 — "auto" also accepted with startingFrameImageUrl
Lengths (s)1, 2, 4, 6, 8, 10, 12, 14, 15
Resolutions360, 540, 720, 1080
Start/end framesstart
Audio outputyes
Max prompt length2024

Kling 3.0 (kling-3.0)

Great with reference start and end frames. Kling's latest and best video model.

Credits4s: 72 · 6s: 108 · 8s: 144 · 10s: 180 · 12s: 216 · 14s: 252 · 15s: 270
Aspect ratios16:9, 9:16, 1:1
Lengths (s)4, 6, 8, 10, 12, 14, 15
Reference imagesup to 3
Reference videosup to 1
Start/end framesstart + end
Audio outputyes
Max prompt length2500

Kling 3.0 Pro (kling-3.0-pro)

Great with reference start and end frames. Kling's latest and best video model.

Credits4s: 96 · 6s: 144 · 8s: 192 · 10s: 240 · 12s: 288 · 14s: 336 · 15s: 360
Aspect ratios16:9, 9:16, 1:1
Lengths (s)4, 6, 8, 10, 12, 14, 15
Reference imagesup to 3
Reference videosup to 1
Start/end framesstart + end
Audio outputyes
Max prompt length2500

Gemini Omni Flash (gemini-omni-flash)

Google's fast video model with sound included. Generate from text, or attach a video and describe the changes you want.

Credits3s: 80 · 4s: 100 · 5s: 130 · 6s: 150 · 7s: 180 · 8s: 200 · 9s: 230 · 10s: 250
Aspect ratios16:9, 9:16
Lengths (s)3, 4, 5, 6, 7, 8, 9, 10
Reference videosup to 1
Start/end frames
Audio outputyes

Veo3.1 Quality (veo3-quality)

Google's highest quality video model with sound included. Blocks celebrities.

Credits8s: 216
Aspect ratios16:9, 9:16
Lengths (s)8
Reference imagesup to 3
Start/end framesstart + end
Audio outputyes

Veo3.1 (veo3-fast)

Google's latest video model with sound included. Blocks celebrities.

Credits8s: 80
Aspect ratios16:9, 9:16
Lengths (s)8
Reference imagesup to 3
Start/end framesstart + end
Audio outputyes

Sora 2 (sora-2)

OpenAI's new bleeding-edge video model with sound included.

Credits4s: 48 · 8s: 96 · 12s: 144 · 16s: 192 · 20s: 240
Aspect ratios16:9, 9:16
Lengths (s)4, 8, 12, 16, 20
Start/end framesstart
Audio outputyes

Sora 2 Pro (sora-2-pro)

OpenAI's new bleeding-edge video model with sound included.

Credits4s: 96 · 8s: 192 · 12s: 288 · 16s: 384 · 20s: 480 (at 1080p)
Aspect ratios16:9, 9:16
Lengths (s)4, 8, 12, 16, 20
Resolutions720, 1080
Start/end framesstart
Audio outputyes

On this page