Logo
Model
Prompt (Required)
Use the formula: Subject + Style + Lighting + Camera + Mood for best results.
0/5000
Resolution
Video Length
15 credits6s
Aspect Ratio
ModeGeneration style mode
Credits required:
15 Credits

Sign in to generate using credits. Video downloads require a paid plan.

Grok Imagine VideoSample clips
1 / 5

Anime cyberpunk skyline

Original prompt: Japanese anime style, cyberpunk full technica, night view of academy city, futuristic giant buildings and neon-lit streetscape, high-tech academy city atmosphere, a silver dragon flying freely in the wind turning into clouds

New · Gemini Omni 1.2 Flash

Gemini Omni 1.2 Flash AI Video Generator

Fast unified video generation up to 4K with native sound, first/last frames, and up to 10 reference images in your browser

Gemini Omni 1.2 Flash combines speed and high fidelity in one engine. Generate 3–15 second clips from text prompts, animate first/last frames, or steer scenes with up to 10 reference images across 360p, 720p, 1080p, and 4K. Native audio is built in, with transparent credit pricing before you generate.

How It Works

Create a Gemini Omni 1.2 Flash Video in 4 Steps

1

Choose Text or Image Input

Start in Text to Video with a prompt, or switch to Image to Video for First / Last Frame animation or up to 10 Reference Images.

2

Describe Motion, Style and Audio

Write the action, camera move, lighting, and sound cues, and optionally set a System Prompt or Negative Prompt to steer tone.

3

Pick Resolution, Ratio and Length

Select 360p, 720p, 1080p, or 4K in 16:9 or 9:16, and set 3–15 seconds (3–10 seconds when locking both first and last frames).

4

Generate and Download

Review the live credit cost, submit your job, and preview or download the MP4 clip directly from your results panel.

Gemini Omni 1.2 Flash Features

4K Clarity, Fast Iteration and Multi-Image Guidance

Four Quality Tiers Including True 4K

Draft in 360p at roughly 1.34 credits per second, render 720p or 1080p at just 2 credits per second, or deliver 4K video — all selectable in one click.

1080p at the Same Price as 720p

With transparent per-second billing matching upstream rates, both 720p and 1080p cost exactly 2 credits per second (20 credits for 10 seconds), letting you target Full HD without a markup.

Text, First Frame or First + Last Frame

Generate purely from text, animate an opening frame up to 15 seconds, or pair a first and last frame (up to 10 seconds) so the clip lands cleanly on your target composition.

Up to 10 Reference Images

Switch Image Use to Reference Images to upload up to 10 images guiding characters, products, costumes, or sets in the same scene.

Native Synchronized Audio

Generates 24 fps H.264 MP4 video paired with matching AAC ambience, sound effects, or spoken cues in one pass (occasional 4K renders may omit audio).

System & Negative Prompt Controls

Fine-tune visual style, camera rules, or excluded elements with dedicated System Prompt and Negative Prompt fields in text, first-frame, and reference modes.

HOW TO USE GEMINI OMNI 1.2 FLASH

How to Make Gemini Omni 1.2 Flash Videos Online

Run the Gemini Omni 1.2 Flash video engine directly in your browser without configuring API keys or cloud quotas. Follow these steps from concept to 4K output.

  1. 1

    Start in the Pre-Selected Playground

    Gemini Omni 1.2 Flash is already loaded above. Sign in with a paid plan or purchased credits when you are ready to render.

  2. 2

    Pick Frames or Reference Images

    In Image to Video, choose First / Last Frame to interpolate between start and end shots, or Reference Images to supply up to 10 visual cues.

  3. 3

    Write Focused Prompts & Constraints

    Keep prompts clear on action and sound, and use System Prompt or Negative Prompt (when not using both start and end frames) to avoid artifacts.

  4. 4

    Choose 16:9 or 9:16 and Quality

    Test composition fast in 360p, publish social clips cleanly in 1080p (only 2 credits/sec), or upscale final showcase renders to 4K.

  5. 5

    Review Audio & Retry Safely

    Check the rendered video and sound sync. Failed upstream tasks are automatically refunded in full to your credit balance.

Pro Tips for Gemini Omni 1.2 Flash

Prototype in 360p, Publish in 1080p or 4K

Use 360p for low-cost motion tests, then switch straight to 1080p (same price as 720p) or 4K once timing looks right.

Respect the 10s First+Last Frame Window

Pairing both first and last frames supports 3–10 seconds and disables system/negative prompts upstream, so put all key guidance in your main prompt.

Direct Sound Explicitly

Mention dialogue, ambient textures, and SFX directly in the scene description so native AAC audio aligns with visual beats.

Use Multiple Reference Angles

Providing consistent front and side reference images among your 10 slots improves character and product stability.

Who Uses Gemini Omni 1.2 Flash

Built for Rapid Iteration and High-Resolution Delivery

Short-Form Video & Reels Teams

Produce crisp 9:16 1080p clips with synchronized audio for TikTok, Reels, and Shorts at just 2 credits per second.

Ecommerce & Product Demos

Combine first/last frame transitions with up to 10 multi-angle product references to keep logos and packaging consistent.

Concept Artists & Motion Designers

Steer mood, palette, and framing rules cleanly using System Prompt and Negative Prompt controls.

Commercial & 4K Showcases

Deliver crisp 4K 16:9 widescreen hero shots directly from still keyframes or descriptive prompts.

Gemini Omni 1.2 Flash FAQ

Everything About Gemini Omni 1.2 Flash

What is Gemini Omni 1.2 Flash?

Gemini Omni 1.2 Flash (gemini-omni-1.2-flash-ext) is a unified multimodal video model supporting text-to-video, first-frame and first+last-frame animation, and multi-image references up to 4K resolution with built-in audio.

What resolutions and durations are supported?

It supports 360p, 720p, 1080p, and 4K resolutions in 16:9 or 9:16 aspect ratios. Text-to-video, single first-frame, and reference modes support 3–15 seconds; combining both first and last frames supports 3–10 seconds.

How many credits does Gemini Omni 1.2 Flash cost?

Pricing is billed per second by resolution: 360p costs ~1.34 credits/s (5s = 7, 10s = 14, 15s = 20), 720p and 1080p cost 2 credits/s each (10s = 20, 15s = 30), and 4K costs ~4.67 credits/s (5s = 24, 10s = 47, 15s = 70). Exact credits display live before submission.

Why do 720p and 1080p cost the same credits?

The upstream provider bills 720p and 1080p at the exact same per-second rate on Gemini Omni 1.2 Flash, so we pass that benefit through directly — 1080p costs the exact same 2 credits/sec as 720p.

How many reference images can I upload?

Up to 10 reference images when Image Use is set to Reference Images. In First / Last Frame mode you can upload 1 image (first frame) or 2 images (first and last frames).

Why are System Prompt and Negative Prompt disabled when using two frames?

The upstream Gemini Omni 1.2 Flash API does not accept negative_prompt or system_prompt when both first_frame_image and last_frame_image are provided, and caps duration at 10 seconds. Our form enforces this upfront so your request is never rejected unexpectedly.

Does Gemini Omni 1.2 Flash generate audio?

Yes. Synchronized AAC audio is generated natively alongside the 24 fps H.264 MP4 video (note that occasional 4K outputs may arrive without an audio track).

Why are only 16:9 and 9:16 aspect ratios offered?

Gemini Omni 1.2 Flash strictly supports 16:9 landscape and 9:16 portrait outputs upstream, ensuring every submission matches the model contract cleanly.

How long does generation take?

Typical generation times range from under a minute for 360p drafts, around 1.5–2 minutes for 1080p clips, and roughly 4–5 minutes for 4K 15-second renders.

What happens if a task fails?

Any task that fails upstream is automatically refunded in full back to your account credit balance.

Can I use generated videos commercially?

Yes, paid subscription plans include commercial usage rights, subject to clearing any third-party brands, likenesses, or copyrighted assets in your inputs.

Are webhooks supported for status updates?

Yes. Every submission attaches an APIMart webhook callback alongside real-time polling so completed renders appear in your history immediately.