GMI Cloud Watch posted an update
Utopai X, a text-to-video model with audio, scored 1,150 on the Artificial Analysis leaderboard, placing second in results dated 29 September. GMI Cloud says the model was trained and is served on its GPUs.
Why it mattersThe model is a post-trained version of MiniMax H3, which scored 1,136 in the same results. Utopai says it used production material and feedback from filmmakers to tune the model, a practical example of a specialist studio adapting a foundation model to its own work. The lead over the third-ranked model is only four points, and GMI says the top models’ confidence intervals overlap. So: notable, but hardly a settled league table.
Discuss: Does fine-tuning on real production feedback look like a meaningful edge for video models, or should we wait for a wider gap?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.