Seedance 2.5、30秒の動画と音声を1回のパスで生成
bytedanceが、ネイティブ音声付きの30秒動画を生成するモデル「seedance 2.5」をリリース。google geminiの10秒制限を上回る。これによりAIが本格的な商用制作ワークフローへ一歩近づいた。
bytedance released seedance 2.5, an upgrade to its video generation model that now produces 30-second clips with integrated audio, exceeding the capabilities of google's gemini omni flash. the model accepts up to 30 images, 10 video clips, and 10 audio files as reference, enabling detailed control over characters, lighting, and camera angles. a demo short film, 'the missing pair', was produced entirely within the model's pipeline. the tool is currently available on jimeng ai and doubao pro, with api access coming later via byteplus modelark.