Developer docs

Generate videos

Text-to-video and image-to-video, durations, resolutions and audio.

Video uses the same endpoint as images — pick a video model.

text

POST /v1/generations

json

{
  "model": "gemini-omni-flash",
  "prompt": "Slow dolly-in on a steaming bowl of ramen, neon reflections, cinematic",
  "aspectRatio": "16:9",
  "resolution": "720p",
  "duration": 8,
  "audio": true
}

Field	Notes
`duration`	Seconds. Must be one of the model's `durations`. Price scales with duration.
`resolution`	`480p` / `720p` / `1080p` where the model lists them.
`audio`	Generate sound (only models with `capabilities.audio`). Some models charge more with audio.
`inputImages`	Exactly one image turns the request into image-to-video (first frame).
`negativePrompt`	Supported by the uncensored `wan-*` models.

Video jobs take 1–6 minutes. Use webhookUrl or poll GET /v1/generations/{id} every few seconds; progress (0–100) is updated while rendering.

Image-to-video

bash

curl https://api.poremu.ai/v1/generations \
  -H "Authorization: Bearer $POREMU_API_KEY" -H "Content-Type: application/json" \
  -d '{
    "model": "kling-3.0",
    "prompt": "She turns and smiles, hair moving in the wind, handheld camera",
    "inputImages": ["https://cdn.poremu.ai/u/…/0.png"],
    "duration": 5,
    "audio": false,
    "webhookUrl": "https://example.com/hooks/poremu"
  }'

Output

json

"outputs": [
  {
    "kind": "VIDEO",
    "url": "https://cdn.poremu.ai/u/…/0.mp4",
    "thumbnailUrl": "https://cdn.poremu.ai/u/…/0.thumb.webp",
    "width": 1280,
    "height": 720,
    "durationMs": 8000,
    "mimeType": "video/mp4"
  }
]

Videos are H.264 MP4. Keep your own copy if you need it beyond the lifetime of your library.