Apple researchers unveil ‘Keyframer’: An AI tool that animates still images using LLMs

Key Points:

  • Keyframer is a new AI tool that uses large language models to animate static images through natural language prompts.
  • Users can generate CSS animation code from a static SVG image and prompt, allowing for easy creation and refinement of animations.
  • Keyframer combines prompting with editing to democratize design, providing accessible prototyping while maintaining user agency.



Apple researchers have introduced a groundbreaking AI tool called Keyframer, leveraging large language models to animate static images through natural language prompts. This innovation, outlined in a research paper on, marks a significant advancement in incorporating AI into the creative process, potentially influencing upcoming Apple products like the iPad Pro and Vision Pro.


Keyframer, detailed in the research paper titled “Empowering Animation Design using Large Language Models,” breaks new ground in applying large language models to animation, addressing challenges such as effectively conveying motion through natural language. The tool allows animators to bring their creations to life seamlessly with just a few simple text prompts, streamlining the animation process that previously demanded extensive manual labor.


Powered by a large language model like GPT-4, Keyframer generates CSS animation code based on a static SVG image and user input. Users can upload an image, enter a text prompt describing the desired animation, and let Keyframer generate the corresponding code. They can then refine the animation by directly editing the CSS code or adding more prompts in natural language, enabling a user-centered approach to design exploration and modification.


Keyframer’s user-friendly features were informed by feedback from professional animators and designers, emphasizing iterative design and creativity. Users found the tool significantly faster and more efficient than traditional methods, allowing them to refine their designs incrementally by adding prompts to animate individual elements separately. This approach empowers users to maintain creative control while harnessing the tool’s generative capabilities, striking a balance between accessibility and user agency.


While AI animation tools like Keyframer have the potential to democratize design, concerns about creative control and satisfaction have been acknowledged and addressed through a combination of prompting and editing features. By fostering a collaborative environment between creators and AI, Keyframer aims to revolutionize animation design by expedite prototyping while preserving user autonomy.


Keyframer’s broader impact extends beyond mere technological advancement, signaling a cultural shift towards AI integration in the creative process. By enabling a more intuitive interaction between humans and technology, Keyframer paves the way for a future where AI plays an integral role in creative endeavors, blurring the lines between creator and creation. Apple’s Keyframer represents a significant step towards reimagining the digital landscape and underscores AI’s potential to reshape the creative industry as we know it.


In a move that could redefine the relationship between technology and creativity, Keyframer embodies the fusion of human ingenuity with AI capabilities, setting the stage for a new era where innovation thrives through collaborative efforts between humans and machines.



Prompt Engineering Guides



©2024 The Horizon