HiDream-O1-Video-1.0 Climbs to No. 7 on Arena Image-to-Video Leaderboard

HiDream.ai's HiDream-O1-Video-1.0 has entered the Arena Image-to-Video leaderboard at No. 7 with a 1456 +/- 8 preliminary preference score, putting the new model within single digits of Google's gemini-omni-flash and close to ByteDance's Dreamina Seedance models. The result gives the Beijing-based generative AI company a visible foothold in a fast-moving video-model race where first-frame fidelity, motion plausibility and prompt following are becoming core measures of product quality.