Is Utopai X on Sume? It is built on MiniMax H3, which Sume lists
Utopai X runs inside Utopai's PAI platform and is post-trained on MiniMax H3. Sume does not list Utopai X; it lists minimax-h3 and minimax-h3-max.

No. Sume does not list Utopai X. Utopai X is a video model that Utopai Studios runs inside its own PAI production platform, and the studio says it is post-trained on MiniMax H3. Sume does list MiniMax H3 and MiniMax H3 Max, so those are the nearest models you can call today.
This post separates what the vendor states about Utopai X from what Sume's video catalog actually offers.
What Utopai says about Utopai X
The facts below come from Utopai's own Utopai X page, read 2026-10-10. The page does not state durations, resolutions or prices, so this post does not either.
| Item | What the page says |
|---|---|
| What it is | The video generation model inside PAI, Utopai Studios' production intelligence platform |
| Base model | Post-trained on MiniMax H3 |
| Access | Through PAI; the page links to PAI and does not describe a standalone API |
| Ranking cited | No. 2 overall on Artificial Analysis' Text-to-Video Leaderboard With Audio as of October 1, 2026 |
| Strengths cited | Audio synchronization and physics |
| Duration, resolution, price | Not stated on the page |
What Sume lists instead
Sume's video generation docs list two MiniMax H3 ids. Both are reachable through the same asynchronous POST /v1/videos job. They are the base family Utopai says it started from, not Utopai's post-trained model, so output will differ.
| Sume id | Duration | Resolution | Notes from the docs |
|---|---|---|---|
| minimax-h3 | 5 to 15 s | Native 480p and 768p | Accepts audio and video references |
| minimax-h3-max | 5 to 15 s | 480p, 768p, 1080p | Faster 768p variant; 1080p is a latent refinement from native 768p; text-to-video, first/last-frame image-to-video, reference-to-video; native stereo audio |
How to try the H3 family today
Read the live catalog first, because limits differ per model, then submit a job. The request shape below is the one in the Sume docs.
curl "https://api.sume.com/v1/videos/models" \
-H "Authorization: Bearer $SUME_API_KEY"
curl -X POST "https://api.sume.com/v1/videos" \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax-h3-max",
"prompt": "A rain-soaked street at night, neon reflecting on wet asphalt",
"duration": 8,
"resolution": "768p"
}'What you cannot get on Sume
If you need Utopai X specifically, use PAI. If you want the MiniMax H3 base through an API with job ids, webhooks and one wallet, the two rows above are the ones to test on your own prompts.
- Utopai X itself: it is not a Sume catalog id, so no Sume call reaches it.
- PAI's project features such as script and earlier-take memory: those belong to Utopai's platform; Sume's docs do not describe an equivalent.
- A guarantee that H3 output matches Utopai X quality: the vendor ranking applies to Utopai X, not to the base model.
Sources
Related posts
More in Models
- Veo 3.1 4K needs an 8-second clip; Sume's Veo rows stop at 720p
Google offers Veo 3.1 1080p and 4K only at 8 seconds. Sume's Veo rows run 720p at 4, 6 or 8 seconds. See the gap and which Sume row reaches 4K.
- Veo 3.1 Lite (Preview) on Sume: what the preview label means
Sume's catalog names the row 'Veo 3.1 Lite (Preview)'. Its limits, the audio rates, and how it differs from the Gemini API's Lite lines for 720p and 1080p.
- Which Sume video row takes 1:1, 3:4 or 21:9? Aspect ratio matrix
Seedance and MiniMax rows take 21:9; Wan 3.0 skips it; Kling 3 stops at 16:9, 9:16 and 1:1; Omni and Veo are 16:9 or 9:16 only. The full matrix by row.
- An OpenRouter-compatible video API: sume/auto or a pinned model
Sume's POST /v1/videos follows OpenRouter's video generation API field for field. Let sume/auto pick the model, or pin a catalog id like seedance-2.5.
Written by Sume