Gemini Omni 1.2 Flash Reference-to-Video API with Image and Video References

Key Features of Gemini Omni 1.2 Flash Reference-to-Video API

• Visual Reference Guidance: Use character, product, or style imagery to guide the appearance of a new scene.

• Motion Reference Guidance: Use short video clips to communicate movement alongside the written scene direction.

• Combined Image and Video References: Bring still imagery and motion examples into the same brief to guide a new generation.

• Prompt-Guided Generation: Describe the intended scene and explain how the supplied materials relate to it.

How to Use Gemini Omni 1.2 Flash Reference-to-Video API on Best Image AI

• Input: Provide a prompt and at least one reference image or video. Either reference type can be used alone, or they can be combined in the same request.

• Reference Preparation: Keep image and video quantities within their respective limits, and use short reference clips. See the API documentation for current asset-count and reference-video duration limits.

• Output Settings: Choose landscape or portrait framing, resolution, and a duration within the supported generation range. The reference clip's length does not set the output length.

• Output: Retrieve the newly generated video after processing completes.

• Billing: Usage is billed per second of requested output duration at the selected resolution; reference videos do not incur an additional charge.

Best Use Cases for Gemini Omni 1.2 Flash Reference-to-Video API Integration

• Visual Pitch Development: Combine a scene brief with selected images or clips to develop a video concept for review.

• Campaign Asset Reuse: Bring existing campaign imagery into the brief for a new video scene.

• Reference Board Workflows: Let creative teams select materials from an asset library and submit them with a generation prompt.

• Motion Concept Exploration: Pair short video references with written direction to explore a new sequence rather than revise the reference clip itself.

Note Reference material is required. Explain the role of each asset in the prompt and distinguish the desired output from the supplied examples. Check reference-video length separately from the requested generation duration.

Gemini Omni 1.2 Flash Reference-to-Video: Comparative Analysis

• Compared with Text-to-Video: Reference-to-Video requires visual material in addition to a prompt. Text-to-Video serves briefs that do not need supplied images or clips.

• Compared with Image-to-Video: Image-to-Video uses an image as the starting frame and optionally another as the ending frame. Reference-to-Video accepts images and short videos as reference inputs for a new scene.

• Compared with Video Edit: Video Edit revises a source video and derives output timing and proportions from it. Reference-to-Video lets the caller select the new clip's duration and orientation independently of the reference footage.