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
| Field | Type | Required | Description |
|---|---|---|---|
prompt | string | yes | What to generate. |
aspectRatio | string | no | One of the model's aspect ratios, e.g. "16:9". |
endingFrameImageUrl | string | no | Image URL used as the last frame. External URLs are auto-imported as assets first. |
referenceImages | {docId, url}[] | no | Yapper image assets. Use the id + url from /assets, a process output, or /assets/import — passing docId preserves asset lineage. |
referenceVideos | {docId, url}[] | no | Yapper video assets (same docId + url shape). |
resolution | number | no | Output resolution height, e.g. 1080. |
startingFrameImageUrl | string | no | Image URL used as the first frame. External URLs are auto-imported as assets first. |
videoLength | number | no | Duration 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).
| Model | Company | Credits | Lengths (s) | Max ref images | Frames | Audio |
|---|---|---|---|---|---|---|
seedance-2.0 | bytedance | 200–760 | 4–15 | 9 | — | yes |
grok-imagine | xai | 58–135 | 4, 6, 8, 10, 12, 14, 15 | — | start | yes |
grok-imagine-v1.5 | xai | 20–260 | 1–15 | — | start | yes |
happy-horse | alibaba | 90–450 | 3–15 | 5 | — | yes |
seedance-2.0-fast | bytedance | 180–680 | 4–15 | 9 | — | yes |
wan-2.7 | alibaba | 30–150 | 2, 4, 6, 8, 10 | 5 | — | yes |
pixverse-v6 | pixverse | 20–370 | 1, 2, 4, 6, 8, 10, 12, 14, 15 | — | start | yes |
kling-3.0 | kuaishou | 72–270 | 4, 6, 8, 10, 12, 14, 15 | 3 | start + end | yes |
kling-3.0-pro | kuaishou | 96–360 | 4, 6, 8, 10, 12, 14, 15 | 3 | start + end | yes |
gemini-omni-flash | 80–250 | 3–10 | — | — | yes | |
veo3-quality | 216 | 8 | 3 | start + end | yes | |
veo3-fast | 80 | 8 | 3 | start + end | yes | |
sora-2 | openai | 48–240 | 4, 8, 12, 16, 20 | — | start | yes |
sora-2-pro | openai | 96–480 | 4, 8, 12, 16, 20 | — | start | yes |
Seedance 2.0 (seedance-2.0)
ByteDance's video model with multi-modal input, lip sync, and multi-shot narrative.
| Credits | 4s: 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 ratios | 16:9, 4:3, 1:1, 3:4, 9:16, 21:9 |
| Lengths (s) | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
| Resolutions | 480, 720, 1080, 2160 |
| Reference images | up to 9 |
| Reference videos | up to 3 (15s total) |
| Reference audio | up to 3 (15s total) |
| Audio output | yes |
| Max prompt length | 20000 |
Grok Imagine (grok-imagine)
Grok's best and latest video model with sound included.
| Credits | 4s: 58 · 6s: 72 · 8s: 86 · 10s: 100 · 12s: 114 · 14s: 128 · 15s: 135 (at 720p) |
| Aspect ratios | 16:9, 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 |
| Resolutions | 720 |
| Start/end frames | start |
| Audio output | yes |
| Max prompt length | 4080 |
Grok Imagine 1.5 (grok-imagine-v1.5)
xAI's Grok Imagine 1.5 image-to-video model with audio included.
| Credits | 1s: 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 ratios | 16:9, 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 |
| Resolutions | 480, 720 |
| Start/end frames | start |
| Audio output | yes |
| Max prompt length | 4096 |
Happy Horse (happy-horse)
Alibaba's reference-to-video model with native audio and 720p output.
| Credits | 3s: 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 ratios | 16:9, 9:16, 1:1, 4:3, 3:4 |
| Lengths (s) | 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
| Resolutions | 720 |
| Reference images | up to 5 |
| Audio output | yes |
| Max prompt length | 5000 |
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.
| Credits | 4s: 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 ratios | 16:9, 4:3, 1:1, 3:4, 9:16, 21:9 |
| Lengths (s) | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 |
| Resolutions | 480, 720, 1080, 2160 |
| Reference images | up to 9 |
| Reference videos | up to 3 (15s total) |
| Reference audio | up to 3 (15s total) |
| Audio output | yes |
| Max prompt length | 20000 |
WAN 2.7 (wan-2.7)
Wan's latest reference-driven video model with image and video guidance for better consistency.
| Credits | 2s: 30 · 4s: 60 · 6s: 90 · 8s: 120 · 10s: 150 (at 1080p) |
| Aspect ratios | 16:9, 9:16, 1:1, 4:3, 3:4 |
| Lengths (s) | 2, 4, 6, 8, 10 |
| Resolutions | 720, 1080 |
| Reference images | up to 5 |
| Reference videos | up to 1 |
| Audio output | yes |
| Max prompt length | 5000 |
PixVerse V6 (pixverse-v6)
PixVerse's latest Fal-hosted video model with optional audio and first-frame image support.
| Credits | 1s: 20 · 2s: 50 · 4s: 100 · 6s: 150 · 8s: 200 · 10s: 250 · 12s: 300 · 14s: 350 · 15s: 370 (at 1080p) |
| Aspect ratios | 16:9, 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 |
| Resolutions | 360, 540, 720, 1080 |
| Start/end frames | start |
| Audio output | yes |
| Max prompt length | 2024 |
Kling 3.0 (kling-3.0)
Great with reference start and end frames. Kling's latest and best video model.
| Credits | 4s: 72 · 6s: 108 · 8s: 144 · 10s: 180 · 12s: 216 · 14s: 252 · 15s: 270 |
| Aspect ratios | 16:9, 9:16, 1:1 |
| Lengths (s) | 4, 6, 8, 10, 12, 14, 15 |
| Reference images | up to 3 |
| Reference videos | up to 1 |
| Start/end frames | start + end |
| Audio output | yes |
| Max prompt length | 2500 |
Kling 3.0 Pro (kling-3.0-pro)
Great with reference start and end frames. Kling's latest and best video model.
| Credits | 4s: 96 · 6s: 144 · 8s: 192 · 10s: 240 · 12s: 288 · 14s: 336 · 15s: 360 |
| Aspect ratios | 16:9, 9:16, 1:1 |
| Lengths (s) | 4, 6, 8, 10, 12, 14, 15 |
| Reference images | up to 3 |
| Reference videos | up to 1 |
| Start/end frames | start + end |
| Audio output | yes |
| Max prompt length | 2500 |
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.
| Credits | 3s: 80 · 4s: 100 · 5s: 130 · 6s: 150 · 7s: 180 · 8s: 200 · 9s: 230 · 10s: 250 |
| Aspect ratios | 16:9, 9:16 |
| Lengths (s) | 3, 4, 5, 6, 7, 8, 9, 10 |
| Reference videos | up to 1 |
| Start/end frames | — |
| Audio output | yes |
Veo3.1 Quality (veo3-quality)
Google's highest quality video model with sound included. Blocks celebrities.
| Credits | 8s: 216 |
| Aspect ratios | 16:9, 9:16 |
| Lengths (s) | 8 |
| Reference images | up to 3 |
| Start/end frames | start + end |
| Audio output | yes |
Veo3.1 (veo3-fast)
Google's latest video model with sound included. Blocks celebrities.
| Credits | 8s: 80 |
| Aspect ratios | 16:9, 9:16 |
| Lengths (s) | 8 |
| Reference images | up to 3 |
| Start/end frames | start + end |
| Audio output | yes |
Sora 2 (sora-2)
OpenAI's new bleeding-edge video model with sound included.
| Credits | 4s: 48 · 8s: 96 · 12s: 144 · 16s: 192 · 20s: 240 |
| Aspect ratios | 16:9, 9:16 |
| Lengths (s) | 4, 8, 12, 16, 20 |
| Start/end frames | start |
| Audio output | yes |
Sora 2 Pro (sora-2-pro)
OpenAI's new bleeding-edge video model with sound included.
| Credits | 4s: 96 · 8s: 192 · 12s: 288 · 16s: 384 · 20s: 480 (at 1080p) |
| Aspect ratios | 16:9, 9:16 |
| Lengths (s) | 4, 8, 12, 16, 20 |
| Resolutions | 720, 1080 |
| Start/end frames | start |
| Audio output | yes |