HiDream-O1-Video-1.0 ranks #6 on Artificial Analysis leaderboard
Summary
HiDream has launched its new video model, HiDream-O1-Video-1.0, which debuts at #6 on the Artificial Analysis Image to Video with Audio leaderboard. This model, designed with native omnimodal capabilities, generates 1080p videos ranging from 5 to 20 seconds with synchronized audio that varies based on the scene. The introduction of HiDream-O1-Video-1.0 marks its entry into a competitive field, ranking closely with other notable models like Dreamina Seedance 2.0 and Vidu Q4 Preview. The model is available for $5.80 per minute through the HiHarness API and can also be utilized in vivago R1 Studio.
Analysis
HiDream: HiDream is an organization focused on generative AI and native omnimodal world models. It released HiDream-O1-Video-1.0 as its native omnimodal video model built for physical consistency and synchronized audio generation. The model is positioned as a new entrant in competitive AI video benchmarks. HiDream-O1-Video-1.0: HiDream-O1-Video-1.0 is a video generation model from HiDream that produces 1080p videos ranging from 5 to 20 seconds with natively integrated audio. It emphasizes scene-determined duration and physical consistency in outputs. The model is accessible via the HiHarness API and vivago R1 Studio and ranks on the Artificial Analysis Image to Video with Audio leaderboard. API Availability: The model can be accessed through the HiHarness API and integrated into vivago R1 Studio. Recent Model Launch: HiDream has introduced a new native omnimodal video model with synchronized audio capabilities. Leaderboard Standing: HiDream-O1-Video-1.0 enters a competitive field of leading AI video models on the Artificial Analysis Image to Video with Audio leaderboard.
Categories
aimachine_learningtechvirtuals