Utopai X ranks second in image-to-video, two Elo points behind the leader

The model, post-trained on MiniMax H3, runs inside Utopai's filmmaking platform PAI and has no public API.

By · Published

Primary source: x.com

Why it matters

Utopai's post-trained model ranks near the top of a public video benchmark, while access remains inside its production platform. The benchmark measures preference; adoption will depend on whether filmmakers find PAI's workflow worth the credit cost and platform restriction.

A camera displays a mountain sunset scene with a hiker, alongside floating panels showing the hiker at different moments; “RW” appears on a rock.

Utopai Studios, led by co-founder and CEO Cecilia Shen, ranks second on Artificial Analysis' image-to-video leaderboard with Utopai X, a model the studio post-trained on MiniMax H3. The October 30th ranking puts Utopai X at 1,150 Elo, two points behind Dreamina Seedance 2.5, in a benchmark that compares human preferences between generated videos.

https://x.com/ArtificialAnlys/status/2115990509944766842…

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

Reader comments

Conversation for this story loads after sign-in.