Gemini Omni Video

Google / Gemini Omni Video — Text to Video (gemini-omni-video)

Gemini Omni Video is a multimodal video generation model that supports text, image, video, audio, and character references for rich, controllable output.

Points forts

• •Multimodal – Combine text, images, video clips, audio, and characters.
• •4K support – Generate up to 4K resolution videos.
• •Character consistency – Reuse character IDs across generations.
• •Flexible duration – 4, 6, 8, or 10 seconds.

Paramètres

• •prompt* – Text description of the video to generate
• •image_urls – Reference images (max 5). Video uses 2 image slots out of 7 total.
• •video_url – Reference video (max 1). Formats: mp4, mov. Uses 2 image slots.
• •video_start – Start timestamp (seconds) for the reference video clip. Defaults to 0.
• •video_end – End timestamp (seconds) for the reference video clip. Defaults to full video length.
• •audio_ids – Audio ID list (max 1). Obtain IDs from the Kie audio library.
• •character_ids – Character ID list (max 3). Each character uses 1 image slot. Obtain IDs from the Kie character library.
• •duration – Video duration in seconds
  • • 4 seconds
  • • 6 seconds
  • • 8 seconds
  • • 10 seconds
• •aspect_ratio – Video aspect ratio
  • • 16:9 (Landscape)
  • • 9:16 (Portrait)
• •resolution – Video resolution
  • • 720p
  • • 1080p
  • • 4K
• •seed – Random seed for reproducibility (0-2147483647)

Tarification

Comment utiliser

1. 1.Write a prompt describing your scene.
2. 2.Optionally attach image, video, or audio references.
3. 3.Choose duration, resolution, and aspect ratio.
4. 4.Submit and download your video.

Autres modèles à essayer