Gemini Omni Flash: Text to Video
Generate cinematic video from natural language prompts using Gemini Omni Flash, transforming text descriptions into a single video output with world-aware motion, lighting, and sound. Ideal for social media content creation, rapid video prototyping, and iterative visual storytelling with conversational editing.
About
This ComfyUI workflow turns plain-language prompts into short, cinematic video using the GeminiVideoOmni node. A Prompt Constructor subgraph (node id 3762e4c1-c5e0-4e1b-9449-e5ea5e9948cd) helps you build a well-formed prompt that explicitly includes duration and aspect ratio—two fields Gemini Omni Video expects in the text. Once composed, the prompt is sent through GeminiVideoOmni to the Gemini API, which synthesizes the clip based on your scene description, camera direction, motion cues, and lighting notes.
What you provide
- Reading the workflow
Runs in ComfyUI. This workflow can't run on this page. Open it in Comfy Cloud with everything loaded, or download it for your own ComfyUI.
Try in CloudQuestions & answers
Yes. Gemini Omni Video expects duration and aspect ratio to be specified in the text. The Prompt Constructor subgraph ensures these fields are included so the request is well-formed.
SaveVideo writes a single video file to your ComfyUI output folder. This workflow focuses on the visual output; it does not handle audio tracks. If the API returns or suggests audio, it is not processed here.
Be concrete and structured: define subject, setting, camera move (e.g., slow dolly-in), lighting (e.g., warm golden-hour rim light), and temporal beats (what happens first/next). Keep duration realistic and aligned with the action to avoid rushed or stretched motion.
First, confirm your API credentials in GeminiVideoOmni. Then verify the prompt explicitly states duration (with units like seconds) and a standard aspect ratio (e.g., 16:9, 9:16, 1:1). Mismatched or missing values are common causes of errors or unexpected framing.