Video generation API

Video Generation

Generate videos using 200+ AI models from a single endpoint. All video models use POST /api/generate with a video modelKey.

Endpoint

POST /api/generate

Video-specific parameters

Parameter	Type	Required	Description
`modelKey`	string	✓	Video model identifier
`prompt`	string	✓	Text description of the video
`inputUrl`	string		Input image URL (for I2V models)
`endImageUrl`	string		End frame image URL (for start-end models)
`aspectRatio`	string		`"16:9"`, `"9:16"`, `"1:1"`
`duration`	number		Video duration in seconds (5, 10, 15)
`quality`	string		`"HD"`, `"1080p"`, `"4K"`
`generateAudio`	boolean		Auto-generate audio for the video
`refVideoUrls`	string[]		Reference video URLs
`refImageUrls`	string[]		Reference image URLs (for ref2vid)

Video generation types

Type	Description	Required params
Text-to-Video (T2V)	Generate video from a text prompt	`prompt`
Image-to-Video (I2V)	Animate a still image	`prompt` + `inputUrl`
Reference-to-Video	Generate video matching reference character/style	`prompt` + `refImageUrls`
Start-End Frame	Generate video between two keyframes	`prompt` + `inputUrl` + `endImageUrl`
Video Edit	Edit an existing video	`prompt` + `refVideoUrls`
Video Extend	Extend a video with more frames	`prompt` + `refVideoUrls`

Quick example

curl -X POST https://api.oyefilmy.com/api/generate \
  -H "Authorization: Bearer of_live_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "modelKey": "seedance_2_0_fast_t2v",
    "prompt": "A drone shot flying over misty mountains at sunrise, cinematic, 4K",
    "aspectRatio": "16:9",
    "duration": 5
  }'

Tip: video generation typically takes 30–120 seconds. Poll GET /api/generate/status/:predictionId every 5 seconds; webhook delivery is not available yet.

Model families

smart_display

Seedance

ByteDance Seedance 2.5, 2.0 (Mini/Standard/Fast), 1.5 Pro — T2V, I2V, Reference-to-Video

25+ variants

smart_display

Kling

Kwaivgi Kling v3.0/v2.6/O3 — Turbo/Pro/4K, avatar, motion control, effects

40+ variants

smart_display

Wan Video

Alibaba Wan 2.7/2.6/2.5/2.2 — T2V, I2V, video extend, character swap

30+ variants

smart_display

Veo / Google

Google Veo 3.1 — Standard, Fast, Lite with T2V and I2V

8 variants

smart_display

Vidu

Vidu Q3/Q2/Q1 — Turbo/Pro/Standard with start-end frame control

25+ variants

smart_display

Hailuo / MiniMax

MiniMax Hailuo 2.3/02 — T2V/I2V/Fast modes

12 variants

smart_display

PixVerse

PixVerse V6/C1 — T2V, I2V, start-end, reference, video extend

10 variants

smart_display

HappyHorse

Alibaba HappyHorse 1.1/1.0 — T2V, I2V, reference, video edit

7 variants

smart_display

Gemini Omni

Google Gemini Omni Flash — T2V, I2V, reference, video edit

7 variants

smart_display

Grok Video

xAI Grok Imagine Video — T2V, I2V, reference, extend, edit

6 variants