Dreamina视频模型展示

Dreamina AI 发布了其最新的视频生成模型 Video 3.0 Pro-Seedance 1.0，在文本到视频（T2V）和图像到视频（I2V）任务中均排名第一。 Video 3.0 Pro 比大多数视频模型快 10 倍以上，使用 NVIDIA-L20 仅需 41.4 秒即可生成 5 秒 1080p 片段。 我一直在测试这个新视频模型——它的指令遵循能力简直不可思议。

Dreamina AI has released its latest video generation model, Video 3.0 Pro-Seedance 1.0, which has ranked the #1 leaderboards in both the text-to-video (T2V) and image-to-video (I2V) tasks. Video 3.0 Pro is over 10x faster than most video models, generating a 5-second 1080p clip in just 41.4 seconds using an NVIDIA-L20. I’ve been testing the new video model — and its instruction-following is unreal.