Video Edit Models
Transform your footage with intelligent editing models. WaveSpeed brings generative AI to the editing timeline. Remove unwanted objects, change video styles, upscale to 4K, and stabilize shaky footage automatically. Post-production tasks that took hours now take seconds.
AI Video Editing Capabilities
Specialized models for every stage of the post-production pipeline.
1. Object Removal (Video Inpainting)
Powered by ProPainter / E2FGVI. Seamlessly remove people, watermarks, or distractions from moving video. Tracks the object across frames and fills the missing area with background details borrowed from other frames (temporal propagation). Best for cleaning up stock footage and removing wires in VFX shots. Also see the Image Eraser for still images.
2. Style Transfer (Video-to-Video)
Powered by Domino / CoDeF. Apply a specific artistic style (e.g., "Van Gogh," "Cyberpunk," "Anime") to live-action footage. Uses a reference image or text prompt to repaint every frame while maintaining the original motion and structure (temporal consistency). Best for music videos, creative ads, and social media trends. Explore more open-source video models.
3. Super Resolution (Upscaling)
Powered by Real-ESRGAN Video / Video2X. Increase video resolution from 720p/1080p to 4K. Hallucinates realistic details and sharpens edges without artifacts. Best for restoring old archives and improving user-generated content quality. Try the free Video Upscaler or the Image Upscaler.
4. Frame Interpolation (Slow Motion)
Powered by RIFE / DAIN. Convert standard 30fps video into smooth 60fps or create fluid slow-motion effects. Generates intermediate frames between existing ones to smooth out motion. Best for sports highlights and cinematic b-roll. Available on WaveSpeed.
Transformation Examples
See the difference AI editing makes.
Q & A
Can AI edit a full video for me?
While current AI excels at specific tasks (like removing an object or changing style), "Auto-Editing" a full movie is still experimental. However, you can chain multiple models on WaveSpeed to automate significant portions of your workflow, such as color grading and cleanup.
Does style transfer flicker?
Early models suffered from flickering. However, WaveSpeed hosts advanced "Consistent Video-to-Video" models (like Domino and CoDeF) that specifically solve this by enforcing temporal consistency, ensuring the style remains stable across frames.
How large of a video file can I upload?
We support uploads up to 2GB for Pro users. For larger files or enterprise workflows, we recommend using our API with cloud storage links (S3/GCS) to avoid browser upload limits.
Is the output compatible with Premiere/DaVinci?
Yes. We export standard MP4 or MOV files (ProRes available on Enterprise plans) that can be dragged directly into any professional Non-Linear Editor (NLE).
Can I automate subtitles?
Yes. Use our "Whisper" model to transcribe audio to text, and then overlay the subtitles onto your video using our rendering API.