Seedance 2.0
Create from text, images, video, and audio references with controllable motion, framing, character, and sound.
CanvasArcCanvasArc AI Video Generator
Create cinematic video from text, images, or an existing clip using production-ready AI models.
CanvasArc AI Video Generator
Video Generator
Your first AI video is free
Continue with Google to generate 1 AI video free — no card required.
0 / 7000
Cost 6 creditsRemaining 0 credits
Video Preview
No Videos Generated
AI VIDEO MODELS
Create with Leading Text-to-Video and Image-to-Video Models
Compare generation modes by input type, reference control, motion quality, and audio support—then select the model inside the studio.
MiniMax H3
Generate 4–15 second 2K videos with native stereo audio from text, first-frame images, or multimodal references.
Veo 3.1
Google's leading video model supports text-to-video, image-to-video, strong prompt alignment, and generated audio.
Sora 2
Designed for controllable, physically coherent video with synchronized dialogue and sound effects on supported workflows.
Kling Video O1 & Wan Pro
Additional options for video-to-video editing and animating a still image through provider-specific modes.
Model availability, duration, resolution, audio, and reference inputs vary by provider.
VIDEO GENERATOR FAQ
AI Video Generation Questions
Which AI video models are available?
The generator offers provider-dependent access to Seedance 2.0, MiniMax H3, Veo 3.1, Sora 2, Wan Pro, and Kling Video O1.
What can MiniMax H3 generate?
MiniMax H3 supports text-to-video, first-frame image-to-video, and multimodal reference workflows. It generates 4–15 second 2K video with native stereo audio and supports prompts up to 7,000 characters.
Can I turn an image into a video?
Yes. Select image-to-video, upload your starting image, and describe the subject motion, camera movement, pacing, and desired atmosphere.
How should I write an AI video prompt?
Describe the shot, subject, action, environment, camera movement, lighting, timing, and audio. For image-to-video, focus primarily on motion and camera behavior.