Alibaba HappyHorse participates in Arena blind testing for the first time and tops the video editing track; the official version will be released in two weeks.

robot
Abstract generation in progress

ME News Report, April 15 (UTC+8), according to Beating Monitoring, Alibaba’s ATH AI Innovation Department’s AI video model HappyHorse announced that version 1.0 has been launched on UC Berkeley community blind testing platform Arena (formerly LMArena). In its first competition, it scored 1299 points (initial score, ±21) to take first place in the video editing track, leading second-place xAI’s grok-imagine-video by 42 points, and third-place Kuaishou KeLing o3-pro by 48 points. HappyHorse has previously topped the rankings on another evaluation platform, Artificial Analysis, in both text-to-video and image-to-video categories, with video editing being its new expansion track. Arena officials state that video editing is an emerging frontier capability of video models, currently supported by only a few models. The HappyHorse team said they are currently in the final optimization sprint before official release, with the official version launching in two weeks. Community users can already experience and vote on arena.ai. (Source: BlockBeats)

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin