Granola raises 125 million dollars at 1.5 billion dollar valuation to turn meeting notes into company wide AI context
March 26, 2026
Mistral AI launches Voxtral Transcribe 2 with real time speech recognition at low cost
March 27, 2026
ByteDance has pushed its latest AI video generation model into the hands of everyday creators through the popular CapCut app. Dreamina Seedance 2.0 lets users turn text prompts, images, audio files or short reference clips into coherent 15 second videos that handle realistic movement and lighting better than before. The rollout starts now for paid users in a handful of countries, with plans to expand later. 📹
Core capabilities of Dreamina Seedance 2.0
The model works with multiple inputs at once. Creators can feed in up to nine images, three video clips and three audio files alongside a text prompt, then direct specific elements with simple tags. It produces clips across six aspect ratios and supports use cases that range from cooking demos and fitness routines to product overviews and action sequences, areas where earlier AI tools often struggled with unnatural physics or jerky motion.
Inside CapCut, the technology appears in features such as AI Video, Video Studio and the broader AI Generator tools. Users can draft ideas quickly, edit existing footage or test concepts before they shoot anything real. The same model already lives in Dreamina and Pippit, and it remains accessible in China through the Jianying app. Details on the official integration appear in the CapCut announcement.
ByteDance built the model on a unified multimodal architecture that improves motion stability and prompt adherence over the previous version. Output includes an invisible watermark to help identify AI generated content once it leaves the platform.
Safety measures and rollout strategy
To address concerns around copyright and deepfakes, Seedance 2.0 blocks generation from real human faces and unauthorized intellectual property. These restrictions likely explain why the initial CapCut release targets markets in Latin America and Southeast Asia rather than the United States or Europe, where Hollywood groups raised issues earlier this year. Coverage from Fast Company notes the gradual global expansion, while The AI Insider highlights the focus on practical creative workflows.
The company says it will keep working with creators and experts to refine the system as more regions come online. In the meantime, the tool gives social media makers and small teams a fast way to prototype videos without full production crews.
Context for AI video creators
Video generation keeps getting more accessible, but quality and control still separate the useful tools from the gimmicks. Dreamina Seedance 2.0 sits alongside other strong options in the space. For a wider look at what is available, see AI video generation tools or dedicated pages on text to video platforms, image to video converters and multimodal video creators.
ByteDance timed the CapCut integration while competitors adjust their own offerings. The result gives millions of editors a direct path to experiment with longer, more stable AI clips inside the app they already use every day.
Will the safety limits and regional rollout slow adoption, or will creators in these first markets show what the model can really do once the guardrails loosen a bit more?
