Gemini Omni Video API – AI Video Generator from Text & Images

What is Gemini Omni Video API?

Gemini Omni Video API is an Apify Actor that generates AI videos from text prompts and reference images using Google Flow and Gemini Omni Flash. Enter a prompt, optionally add an image URL or base64 image, and get a downloadable MP4 video back. Try it for free on Apify — no browser automation, Google login, or setup needed on your side.

This actor works like a simple Gemini Omni API for video generation: submit input, start the run, and collect the generated video from Apify storage.

What can you create with Gemini Omni Video API?

• Text-to-video — describe a scene and generate a short cinematic video
• Image-to-video — add a reference image and animate or transform it
• Scene transformation — change mood, lighting, background, camera angle, or visual style
• Product demos — turn screenshots, product images, or landing pages into short promo clips
• Social media videos — create clips for TikTok, Reels, YouTube Shorts, ads, and storyboards
• Visual storytelling — generate coherent scenes with clear motion, style, and direction

Gemini Omni Flash is Google’s first Omni model for practical video generation workflows, including natural language edits, reference-based creation, and visual storytelling.

Gemini Omni example videos on YouTube

Watch these examples before writing your prompt to see what Gemini Omni Flash can do:

• Official Google launch video: Introducing Gemini Omni: Create Anything from Anything
• Official Google community showcase: See how our community is using Gemini Omni features to create magic
• Creator walkthrough: I Can't Believe I Put Myself in an AI Video in 2 Minutes | Gemini Omni
• Google Flow ad workflow: How to Make UGC Ads With Gemini Omni ($0.60 Per Video)

What does Gemini Omni Video API return?

Field	Description	Example
success	Whether the video generation succeeded	true
prompt	Prompt used for generation	"Create a cinematic drone shot..."
videoKvsKey	Key of the MP4 video in Apify key-value store	"video-1719334800.mp4"
videoKvsUrl	Direct Apify storage URL for the generated video	"https://api.apify.com/..."
model	Model used	"Omni Flash"
hasImage	Whether a reference image was used	true
timeMs	Total processing time in milliseconds	420000
checkTime	ISO timestamp	"2026-06-25T21:30:00.000Z"

How to generate Gemini Omni videos on Apify

1. Open Gemini Omni Video API on Apify Store and click Try for free
2. Enter a detailed prompt describing the video you want
3. Optionally add imageUrl or imageBase64 as a reference image
4. Click Start — the actor waits up to 30 minutes for the video to complete
5. Download the MP4 video from the key-value store using videoKvsKey or videoKvsUrl

Text-to-video input example

{
  "prompt": "A drone shot flying over a neon-lit cyberpunk city at night, rain reflecting on the streets, cinematic, slow camera movement, 8 seconds"
}

Image-to-video input example

{
  "prompt": "Smooth cinematic zoom into this hotel listing screenshot, highlight the booking interface, upbeat product demo style",
  "imageUrl": "https://example.com/hotel-screenshot.jpg"
}

How much does Gemini Omni Video API cost?

You can try the actor for free with Apify platform credits. Pricing is designed around successful video generation.

Component	Cost
Successful video generation	$0.99 per generated video
Apify compute	Usually ~$0.06–$0.13 for a 30-minute run at 512 MB

Typical total cost is about $1.05–$1.12 per successful video. Failed generations are not charged the successful-video event fee; they only use Apify compute for the time spent.

On the Apify Free plan with $5 platform credit, you can test roughly 3–4 video generations, depending on runtime.

Input

Parameter	Type	Required	Default	Description
prompt	string	yes	—	Describe the video: subject, style, camera movement, mood, lighting, and action
imageUrl	string	no	—	Public URL of a reference image
imageBase64	string	no	—	Base64-encoded image data, alternative to imageUrl

See the Input tab for the full schema with examples and prefilled values.

Output example

{
  "success": true,
  "prompt": "Create a cinematic demo video with smooth zooms",
  "error": null,
  "message": "Video generated successfully.",
  "videoKvsKey": "flow-video-1719334800.mp4",
  "videoKvsUrl": "https://api.apify.com/v2/key-value-stores/.../records/flow-video-1719334800.mp4",
  "model": "Omni Flash",
  "hasImage": true,
  "imageUrl": "https://example.com/screenshot.jpg",
  "timeMs": 420000,
  "checkTime": "2026-06-25T21:30:00.000Z"
}

Download the video file from Apify key-value store using videoKvsKey or videoKvsUrl. Export dataset results as JSON, CSV, Excel, HTML, or XML, or pull them via the Apify API.

Tips for better Gemini Omni video prompts

• Describe motion clearly — camera push-in, pan left, handheld, drone shot, slow zoom
• Mention style — cinematic, UGC ad, product demo, documentary, anime, claymation, photorealistic
• Use reference images when you need a specific product, screenshot, layout, or character look
• Keep prompts focused — one clear scene usually works better than many unrelated ideas
• Allow enough time — default timeout is 30 minutes, but the actor uses only 512 MB by default to keep Apify compute low

Can I use Gemini Omni Video API programmatically?

Yes. You can run this actor from the Apify API, Apify SDK, scheduled tasks, webhooks, Make, Zapier, Google Sheets, or your own backend. Each run returns structured dataset output and stores the final MP4 video in Apify key-value storage.

FAQ

Is there a free trial? Yes. Click Try for free on Apify and use your free Apify platform credits to test Gemini Omni video generation.

Do I need a Google account? No. For normal use, you only provide a prompt and optional image. The backend handles the Google Flow session.

What video format do I get? The actor saves the generated video as an MP4 file in Apify key-value storage.

How long does generation take? Most videos finish in 2–10 minutes. The default maximum wait time is 30 minutes so slower generations can still complete.

Can I generate a video without an image? Yes. If you only provide prompt, the actor runs text-to-video generation.

Can I use a reference image? Yes. Use imageUrl for a public image URL or imageBase64 if you already have the image data.

What if generation fails? The result includes success: false, an error code, and a message. You are not charged the successful-video event fee for failed generations.

Is it legal to use Gemini Omni Video API?

This actor generates AI videos using Google Flow and Gemini Omni Flash. Do not use generated content to deceive, impersonate, harass, or harm others. You are responsible for making sure your prompts, reference images, and generated outputs comply with applicable laws, platform rules, and rights of third parties.

---

Google Flow, Google AI, and Gemini Omni are trademarks of Google LLC. This actor is not affiliated with or endorsed by Google.