Omni Flash Video Generation

Omni Flash is a video series focused on speed and cost-effectiveness, suitable for high-volume production and fast-paced content. Requests use the unified asynchronous protocol; see Video Generation (Asynchronous Protocol).

Model Naming

omni_flash[_method][_duration]_aspect_ratio:

• Method: defaults to text-to-video; i2v for image-to-video; r2v for reference image-to-video;
• Duration: 4s / 6s / 8s / 10s (the default model uses the default duration);
• Aspect ratio: portrait (vertical) / landscape (horizontal).

Examples: omni_flash_6s_portrait, omni_flash_i2v_10s_landscape, omni_flash_r2v_8s_portrait.

Submission Example

curl https://api.4allapi.com/v1/video/generations \
  -H "Authorization: Bearer $AIPROXY_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "omni_flash_6s_portrait",
    "prompt": "咖啡拉花特写,奶泡缓缓画出天鹅图案,暖色调"
  }'

Usage Notes

• Just include the duration in the model name; do not pass an extra duration parameter;
• Short durations (4s/6s) render faster and cost less per clip, making them suitable for batch testing before selecting the better ones for longer durations;
• For vertical scenarios such as e-commerce/social media deliverables, use _portrait directly;
• Trade-off with VEO: Omni Flash is faster and more economical, while VEO has stronger visual quality; see VEO 3.1.