Magic Hour, an AI video and image creation platform, today announced updates to its text-to-video and image-to-video products, enabling creators to produce up to 60-second videos using Sora 2 and Veo 3.1 through a longer-form workflow designed for publish-ready outputs.
Many generative video models are optimized for short clips. Magic Hour’s approach generates a longer video by sequencing multiple short generations into a single timeline and applying continuity controls designed to improve visual consistency across segments so the final output reads as one piece of content rather than a collection of disconnected clips.
“Creators don’t want eight or ten seconds. They want a complete idea,” said Runbo Li, co-founder and CEO of Magic Hour. “Short clips are useful, but they break down when you’re trying to tell an actual story. We built this so longer videos can feel coherent from start to finish.”
How It Works (High Level)
- Text-to-video: Users start with a prompt. Magic Hour generates an ordered shot plan, produces the segments, then assembles them into a single video output.
- Image-to-video: Users start with a reference image to anchor key visual elements. Magic Hour generates the segments and assembles the final video while carrying forward selected visual cues across the output.
Why This Matters
Short-form generations are effective for rapid concepts, but longer formats support use cases that require pacing and structure, including product demos, short narrative sequences, marketing creatives, and social content where context matters.
Additional Product Launches Included in This Update
- Z-Image Turbo: Faster image generation for high-throughput iteration
- Qwen Edit: Image editing using the latest Qwen model integration
- Talking Photo: Animate a still image into a speaking video format
- Kling 2.5 Turbo: A fast mode for short-form generation and rapid iteration
Availability
The 60-second workflow is available now in Magic Hour’s text-to-video and image-to-video products. Availability and performance may vary based on model capacity, queueing, and workload complexity.
About Magic Hour
Magic Hour is an AI video creation platform that helps creators and teams generate and edit videos using text and images, with workflows designed for practical outputs and repeatable creative control. Learn more at magichour.ai.
Disclaimer
Sora, Veo, Qwen, and Kling are trademarks or registered trademarks of their respective owners. Magic Hour is not affiliated with, endorsed by, or sponsored by these entities.
Media Contact
Runbo Li
Co-founder and CEO, Magic Hour
press@magichour.ai
