Using starting and ending frames—or reference images—as guides for video generation has become a widely adopted technique in the creative AI community. We’ve observed many creators consistently leveraging this approach in tools like Kling and Google Flow, especially for generating more coherent, visually aligned results.
Today, we’re excited to announce that Crevas now fully supports both keyframe-based and image-referenced text-to-video generation.
You can now define a first frame, a last frame, or provide one or more reference images to guide your video generation with more control and stylistic consistency.
This new feature is designed to help you:
• Set clear visual direction for your story.
• Achieve smoother scene transitions.
• Maintain stylistic alignment throughout your video.
Whether you’re creating cinematic shots, product demos, or abstract concepts—this workflow gives you the precision you need.
Ready to try it? Jump into the canvas and start creating with image guidance today.