wan-2-2/lora/image-to-video
Transform stills into cinematic motion with open-source precision tools.
This model reads one or more reference images together with a written prompt, then generates a short cinematic clip that keeps the referenced subject, look, and composition while adding motion and sound. It belongs to Google's Gemini Omni Flash family, which also covers text-to-video and image-to-video; this RunComfy page runs the reference-to-video task.
Because the pictures anchor identity and visual direction, the same character or product can stay recognizable from the first frame to the last instead of drifting between shots.
Inputs match the RunComfy OpenAPI Input schema for the reference-to-video task.
| Parameter | Required | Type | Default | Range / Options | Description |
|---|---|---|---|---|---|
| prompt* | Yes (*) | string | — | descriptive text | Directs motion, pacing, scene development, and audio |
| image_urls* | Yes (*) | array (image URIs) | — | jpg, jpeg, png, webp | One or more reference images guiding subject, style, and layout |
| aspect_ratio | No | string | 16:9 | 16:9, 9:16 | Aspect ratio of the generated video |
| duration | No | integer | 8 | 3-10 (seconds) | Length of the generated video in whole seconds |
Billing is based on the length of the generated clip:
Transform stills into cinematic motion with open-source precision tools.
Add instant visual effects to a single image and export as a video.
Unified AI model for refined scene editing, style match, and smooth video refits
Generate high quality videos from text prompts with Wan 2.2 Plus.
Build a scene from 1–6 images and animate it into a video.
Convert photos into expressive talking avatars with precise motion and HD detail
Gemini Omni Flash Reference to Video takes one or more reference images and a text prompt, then generates a short clip with synchronized audio. The reference images guide the subject, style, and layout, while the prompt directs motion, pacing, scene development, and sound, so your pictures set the look and the text brings it to life.
It works well for keeping a character, product, or brand look consistent across a short clip, building reference-guided cinematic shots, and producing promos from existing visuals. Marketers, filmmakers, game teams, and social creators often use Gemini Omni Flash Reference to Video when identity and visual direction need to stay close to their source images.
You can provide one or more reference images, and supported formats include jpg, jpeg, png, and webp. With Gemini Omni Flash Reference to Video you can mention what each image contributes in the prompt, such as the character, outfit, or environment, to steer how it shapes the result. Check the current RunComfy parameter panel for the exact input limits.
The model generates sound in step with the video, including speech, ambient effects, and music where appropriate. You can steer this in the prompt with Gemini Omni Flash Reference to Video by naming the dialogue tone or background sound you want, or asking for no dialogue when you need a quiet clip.
Supplying reference images gives you tighter control over subject identity, style, and layout than a text-only prompt. With Gemini Omni Flash Reference to Video, the images anchor how the shot looks while the prompt directs how it moves, which helps when you already have specific characters or scenes you want to carry into the clip.
Gemini Omni Flash Reference to Video supports 16:9 landscape and 9:16 portrait framing, with clip lengths from 3 to 10 whole seconds. Check the current RunComfy parameter panel for the exact options, since available settings can change over time.
Yes. You can prototype Gemini Omni Flash Reference to Video in the RunComfy AI model UI, then call the same model via the RunComfy HTTP API with identical parameters for automation and production. This lets you move from testing to integration without hosting or scaling the model yourself.
Generations with Gemini Omni Flash Reference to Video are billed by clip length at $0.14 per second of generated video, and consume usd or credits on your RunComfy account. New users typically receive a free trial amount, and you can review current rates in the Generation section of this page.
RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.
