PixVerse C1 Image-to-Video

## Overview

PixVerse C1 is a video generation model from PixVerse. It turns a still image plus your instructions into a short MP4 video clip. It uses your image as the starting frame, then generates motion, camera movement, and scene dynamics across time. This helps you turn a keyframe or concept frame into a usable shot for an edit.

## What you can build

- Product shots that gain motion, camera push-ins, and atmosphere from one photo.

- Character or portrait animation clips for mood boards and story reels.

- First-to-last-frame transitions where the clip starts on one image and ends on another.

- Short cinematic cutaways and establishing shots, driven by prompt details.

## Inputs

- A start-frame image you want to animate. Common formats include JPG, JPEG, PNG, WebP, GIF, AVIF, HEIC, and HEIF. If you provide a start image, it acts as the first frame of the video.

- An optional end-frame image to force a specific final frame. When you add this, the model generates a transition from the start image to the end image.

- A written prompt that describes the motion, camera, setting, and actions you want to see. Very long prompts may be rejected; keep it under 2048 characters.

- A clip duration selection: 5, 10, or 15 seconds.

- A quality setting that picks the target output resolution: 360p, 540p, 720p, or 1080p.

- An aspect ratio choice for text-only generation. If you provide a start image, the model derives framing from that image, so aspect ratio controls don't apply.

- An optional random seed number for repeatable results. Keeping the same seed and prompt can reproduce the same video output. Setting the seed to 0 randomizes it.

- An optional toggle to generate audio alongside the video. Audio can include background music, sound effects, and dialogue.

## Outputs

The model returns a generated MP4 video file. If you enable audio, the MP4 includes an audio track.

## Recommended settings

- Start with 540p and a 5-second clip while you refine prompt wording and framing.

- Use a fixed non-zero seed when you want controlled A/B changes between runs. Leave it at 0 when you want new motion each run.

- Add an end-frame image only when you need a clear final pose or composition. Otherwise, let the model choose a natural ending from the prompt.

## Limitations

- Output length is limited to short clips. C1 supports up to 15 seconds.

- When you animate a provided start image, framing is tied to that image. Aspect ratio settings won't override it.

- Prompts have a strict size limit. Very long prompts can fail.

- Fine text inside the scene can drift or deform across frames. Logos and small typography often need post work.

- Low-quality inputs raise failure risk. Blurry images, heavy JPEG artifacts, screenshots of screenshots, or images with tiny subjects can cause warping and identity drift.

- Transition mode can produce flicker if the start and end frames have mismatched lighting, lens style, or viewpoint.

## Safety & compliance

A safety checker is enabled by default. Don't use this model to generate disallowed sexual content, violence, or hateful imagery.

Only upload images you own or have rights to edit. Get clear consent before using real people's faces, especially for ads, political content, or impersonation.