Video Generation Overview
Turn any character into AI video — the studio, the four models, the action library, text-to-video vs image-to-video, settings, and cost.
Video generation turns a character into moving, intimate video. It's one of Xotic AI's headline features and lives in the Create studio's Videos view.
The basic flow#
- Pick a style — Realistic or Anime.
- Choose an action — browse the action library and pick the motion/scene you want. Each action is a preview thumbnail. There are hundreds of them.
- Set your options — frame rate, length, upscaling, and (for some models) a starting image or spoken text.
- Generate — and your video renders.
Two ways to generate: T2V and I2V#
- Text-to-Video (T2V) — generate a video straight from your character and the chosen action. No starting image needed. This is how most actions work.
- Image-to-Video (I2V) — start from a still image and animate it. Some actions are I2V-only (they need a starting image). Because the video grows out of that frame, the starting image is the single most important input — see Text-to-video & image-to-video.
Four models#
Xotic AI has four video models (V1–V4), each with its own action set and strengths: V1 for fast, ~7.5-second clips from a huge action library; V2 for long clips (up to 30 seconds) with sound; V3 for fully custom, do-anything scenes; and V4 for lip-synced talking video. Full breakdown in Models & actions.
Settings#
Depending on the model, you can control:
- Frame rate — 16 FPS (quick), 32 FPS (standard), or 64 FPS (smoothest).
- Length — V1 clips run about 7.5 seconds; V2 goes from 5 up to 30 seconds and includes audio.
- Upscaling — render at higher resolution for extra XOT.
Cost#
Video generation costs XOT, and the amount depends on the model, length, frame rate, and upscaling — roughly from 150 XOT for a short V1 clip up to much more for long, high-frame-rate, upscaled V2 video. The studio always shows the exact cost before you generate. See XOT costs.
Stitching clips together#
Made several clips? The video stitch studio in Create lets you combine them into a longer sequence.
Go deeper#
- Models & actions — the four models and the action library
- Text-to-video & image-to-video — and why the starting image is everything
- Video prompting tips — get dramatically better results