Sign in
Pending benchmarksTracking since Nov 1, 2026

Is Utopai X state of the art?Pending benchmarks.Announced Sep 30, 2026; no independent results yet.

Utopai Studios' video model with audio, made for filmmaking. Utopai, a film and TV studio, post-trained it on MiniMax's MiniMax H3 using its own production data and filmmaker feedback. It launched on September 30, 2026 as text to video inside PAI, Utopai's paid production platform, and added image to video by late October. Use costs 93 PAI credits per second of 1080p video. There is no public API, and Utopai calls the model proprietary.

Announced facts

The details so far

Lab
Utopai Studios
Announced
Sep 30, 2026
API
Not yet
Open weights
Closed · Not released; Utopai calls it its proprietary video model, and it has no repos on Hugging Face
Base model
MiniMax H3, post-trained by Utopai Studios (a derivative; not a MiniMax release). Stated on Utopai's Utopai X page and in its Oct 2 and Oct 29 blogs
Modes
Text to video with audio at launch (Sep 30, 2026); image to video added by Oct 29, 2026
Access
Only inside PAI (Production Assistive Intelligence), Utopai's web platform for film and TV production; no public API
Price
93 PAI credits per second of 1080p video; PAI plans start at $15 a month for 1,000 credits (Utopai's PAI page, Nov 1, 2026)
Training data
Utopai's own production material (scripts, approved designs, shot plans and takes) and filmmakers' judgments about which takes work, per Utopai
Release history
Launched Sep 30, 2026 with the relaunched PAI; the Oct 29-30, 2026 posts announce the image-to-video mode, not a new model

Tags marked unsure are not stated on the lab's own pages.

Next steps

What would let us call it

Verdicts use only independent boards we already track. Two consecutive snapshots are required before any crown.

  • Publicly announced with a source
  • Usable by the public (API or weights)
  • Listed on an independent board we track
  • Enough results for a rank in a category
  • Holds across two snapshots (crowns only)
Self-reported

What the lab claims

Self-reported. Shown for context; not used in any verdict.

Lab claim: No. 2 on Artificial Analysis' text-to-video with audio leaderboard at launch, with an Elo of 1,150 in the Sep 29, 2026 results, and the highest-ranked model from a U.S.-based company there.

Lab claim: No. 2 on Artificial Analysis' image-to-video with audio leaderboard (AA-Video-I2V v2.1) in the Oct 29, 2026 results.

Receipts

Sources

LABUtopai blog: Introducing the All-New PAI and Utopai Xutopaistudios.com/en/blog/introducing-the-all-new-pai-and-utopai-x · Sep 30, 2026LABUtopai Studios launch post on Xx.com/UtopaiStudios/status/2105352264294793639 · Sep 30, 2026PRESSArtificial Analysis on Utopai X's text-to-video debut (X)x.com/ArtificialAnlys/status/2105343643251032304 · Sep 30, 2026LABUtopai X product page and FAQutopaistudios.com/en/pai/utopai-xLABUtopai blog: Utopai X capabilities for filmmakingutopaistudios.com/en/blog/utopai-x-ai-video-model-filmmaking · Oct 2, 2026LABUtopai blog: Utopai X image-to-video with audio resultsutopaistudios.com/en/blog/utopai-x-image-to-video-audio-ranking · Oct 29, 2026PRESSArtificial Analysis on Utopai X image to video (X)x.com/ArtificialAnlys/status/2115990509944766842 · Oct 29, 2026LABUtopai Studios image-to-video post on Xx.com/UtopaiStudios/status/2116022088062603342 · Oct 30, 2026LABPAI plans and credit rates (Utopai)utopaistudios.com/en/pai

Tracking ends Jan 30, 2027 unless the API or weights become available.