Gemini Omni Flash AI Generator
Generated on PixelDojo with Gemini Omni Flash. Produced by PixelDojo's generation pipeline.
Gemini Omni Flash charges 2 credits for every second of video it renders. We ran it at the 3 second minimum the API accepts, so our clip billed 6 credits and came back in about 48 seconds on August 26, 2026, as a 1280 by 720 file carrying a real audio track. This page walks through what that rate card looks like at other lengths, what the model needs as input, and when a still-image edit is the cheaper answer instead.
The 3 second, 6 credit orbit we timed
Every example below was produced on PixelDojo. Hover to see the prompt.
Slow cinematic camera orbit around the earbuds case, soft studio lighting shifts, the lid opens gently to reveal the earbuds, product commercial style
Gemini Omni Flash
Every setting that changes the price or the wait
The price is per second, not per clip
2 credits for every second of output, with no separate charge for aspect ratio or for sending reference images. The app's default duration is 8 seconds, which prices out at 16 credits. We ran the 3 second floor instead, for 6 credits, to see the cheapest the rate card gets.
What the earbuds run measured
We fed it the same product still used elsewhere in this batch and asked for a slow camera orbit with the case lid opening. The finished file was a 1280 by 720 pixel MP4, 3.0 seconds long, 703 KB, with a real audio track: reading it with ffprobe showed an AAC stream averaging -16.7 dB, not a silent placeholder.
Turnaround for our clip
About 48 seconds from submitting the job to the finished file landing in temporary storage, timed on August 26, 2026. The request hands back a job id right away and settles after that, so busier queues or a longer duration will move this number even though the price per second stays fixed.
Three ways to feed it
Text alone for text-to-video, a starting image for image-to-video, or a source video to edit in place. We used the image path. Editing a source video keeps the output at that clip's own dimensions instead of picking one of the two aspect-ratio presets.
720p only, two shapes
16:9 or 9:16 are the only aspect ratios, and both render at 720p. There is no 1080p or 4K setting on this model.
The picker shows four lengths, the API takes ten
The app's duration control offers 4, 6, 8 or 10 seconds. The schema behind it accepts any whole number from 3 to 10, which is how our 3 second test got through even though it is not one of the four buttons.
Continuing a clip locks in the top length
Pass back a previous generation's video id to extend it, and the model regenerates at its standard maximum length rather than your original duration. A continuation is billed at that 10 second maximum so the charge never lands under what actually renders.
Reference images, capped tighter when editing
Up to 7 reference images can guide a text or image generation. Editing an existing video caps that at 5 instead, alongside the source clip itself.
One image, one prompt, one run. 3 seconds of video, 6 credits, about 48 seconds turnaround, with audio in the same file.
Why Choose Pixel Dojo for Gemini Omni Flash
Professional-quality results with cutting-edge AI technology
2 credits per second, no size surcharge
The rate holds regardless of aspect ratio or how many reference images ride along. Seconds times two is the entire formula.
About 48 seconds, measured
Timed from job submission to finished file on August 26, 2026, for a 3 second clip built from one first-frame image and no reference images.
Audio in the same request
No separate voice or sound pass. Our clip's audio track measured real signal on a volume check, not silence, straight out of the one call.
How It Works
How the earbuds clip was made:
Start from a still, not text alone
We passed the earbuds product photo as the first-frame image and let the prompt describe the camera move and the lid opening from there.
Pick a duration, since it sets the whole price
We set 3 seconds, the minimum the API accepts, to see the floor of the 2-credit-per-second rate. It came to 6 credits.
Submit over the API or the app
POST to /api/v1/models/google-gemini-omni-flash/run, or run it from the tool page. Ours returned a 1280 by 720 MP4 with audio in about 48 seconds.
Loved by creators on PixelDojo
Real feedback from people using PixelDojo, pulled from our in-product surveys.
Love how I can almost create anything
The Flux Pro Ultra is just amazing!
Very easy to use, and they have fast wan 2.2 video generation with custom loras available.
The overall quality of the site and it's amazing variety of tools. The continued updates of the UI. The unbelievable level of tech support.
THIS IS SO DOPE !
I have already recommended it to friends
Explore more AI tools on PixelDojo
AI Tools
Compare & Switch
- Best AI Image Generators
- Best AI Video Generators
- Midjourney Alternatives
- Civitai Alternatives
- Runway Alternatives
- Leonardo Alternatives
- Pika Alternatives
- Luma Alternatives
- Magnific Alternatives
- Veo Alternatives
- Flux Alternatives
- Freepik Alternatives
- Seedance Alternatives
- Seedream Alternatives
- Pixverse Alternatives
- GPT Image Alternatives
- Synthesia Alternatives
- Playground Alternatives
- NightCafe Alternatives
- Canva AI Alternatives
- ElevenLabs Alternatives
- ComfyUI Alternatives
- Fal Alternatives
- Replicate Alternatives
Common Questions
Everything you need to know about Gemini Omni Flash
What does Gemini Omni Flash cost?
2 credits per second of output, at any whole-number duration from 3 to 10 seconds. The app's default of 8 seconds comes to 16 credits. Our test ran the 3 second minimum for 6 credits.
How fast is Gemini Omni Flash?
About 48 seconds for our 3 second clip, timed from submission to finished file on August 26, 2026. Longer clips and busier queue load will move that number. The credit rate does not move with it.
Does Gemini Omni Flash generate audio?
Yes, in the same request that renders the picture. Our test clip's audio track measured real signal on a volume check rather than a silent track, with no separate sound step.
How do I call it from the API?
POST to /api/v1/models/google-gemini-omni-flash/run with your API key, a prompt, and an image URL if you want image-to-video. It is the same key and the same request shape as every other model in the catalog.
When should I use Nano Banana Edit instead?
When the job only needs the still edited, not turned into motion. Nano Banana Edit charges a flat 3 credits per image for that and does not touch video at all, so it is the cheaper and quicker pick when a moving camera and native audio are not part of the brief.
Can it edit a video I already have instead of generating a new one?
Yes. Send a source video URL with an edit instruction instead of an image, and the output matches that source clip's own dimensions. Reference images are capped at 5 in that mode instead of the usual 7.