Gemini Omni Flash is Google's video generation model, available in Runway. Use it to generate new videos from a text prompt and optional reference images, or to edit existing footage using a text prompt and optional reference images. 

This guide covers what the model can do and how to generate with it.

Step 1 – Selecting Gemini Omni Flash

The model picker lives at the bottom of the generation panel in Tool Mode. Gemini Omni Flash is not available in Explore credit mode.

1. From the left sidebar, click Custom.
2. Select the Video tab at the top of the generation panel.
3. Click the model dropdown at the bottom of the panel and choose Gemini Omni Flash.

Step 2 – Adding your prompt and references

Gemini Omni Flash has two generation modes, toggled at the top of the model panel.

Frame

Generate a new video from a text prompt and optional first frame or reference images.

1. Type your prompt in the text field.
2. Set Duration using the slider — choose between 3 and 10 seconds.
3. To add an image to act as the first frame, upload or drag and drop to the First Video Frame area.
4. To add reference images, click one of the Reference slots above the prompt field. You can add up to 10 images.

The optional First Video Frame becomes the literal first frame of the generated video. References, by contrast, guide the look and feel of the output without appearing as an exact frame.

You can combine these inputs however you like: a first frame and references together, a first frame alone, or references alone.

Edit Video

Make adjustments to an existing video using a text prompt and optional reference images.

1. Upload the video you want to edit. The file must be under 100 MB and no longer than 10 seconds.
2. Type your prompt describing the change you want to make.
3. Optionally, add a reference image to guide the edit.

Step 3 – Generating and understanding your output

Click Generate to start the job.

Gemini Omni Flash determines the aspect ratio automatically based on your prompt and any reference inputs — there are no manual controls for this setting. The model will output at either 16:9 or 9:16, whichever best fits your inputs.

To influence the aspect ratio, specify your preference directly in your prompt. For example:

• "Generate a 16:9 video of..."
• "Create a vertical clip of..."

All videos are output at 720p resolution.

Example prompts and outputs

Mode	Input (First frame/Video)	Prompt	Reference image(s)	Output
Frames		Camera fluidly pushes into the driver's side window where a man wearing the glasses from @Image 1 is driving the car with snowy hills in the background. Single, continuous shot
Frames	N/A	the woman from @Image 2 wears the glasses from @Image 1 while holding the soda from @Image 3 for a high-fashion product ad for the soda.
Edit		Add fire to the palms of their hands that disappears when their palms face downwards. Make the room dark.	N/A
Edit		change her outfit to match the sequined outfit in @Image 1

Next Steps

You now have a generated video ready to download or pull into your next project. From here, you might want to:

• Upscale your video to 4K — since Gemini Omni Flash outputs at 720p, upscaling is a natural next step before sharing or publishing
• Assemble your clips in Studio — trim and arrange multiple generations into a longer, cohesive video
• Continue in Agent — input your newly generated videos and collaborate with Runway Agent to develop your idea further across multiple tools and outputs

Specs

FAQ