MiniMax-Hailuo-2.3-Fast: an image-to-video model id, no Sume twin
MiniMax lists MiniMax-Hailuo-2.3-Fast in its image-to-video API but not in its text-to-video enum. What that means, and what Sume offers for first frames.
MiniMax-Hailuo-2.3-Fast appears in MiniMax's image-to-video model list, while the text-to-video reference lists only Hailuo 2.3, Hailuo 02 and two older T2V-01 ids (read 2026-10-02). So a text prompt alone cannot call the Fast variant on those pages: it starts from a still. Sume has no Hailuo 2.3 Fast id; for a still-first workflow use minimax-h3 or minimax-h3-max with a first_frame image.
Which model ids does each page list?
The two references enumerate different sets, so check which endpoint you are calling before choosing an id.
| Endpoint | Model ids listed |
|---|---|
| Text-to-video | MiniMax-Hailuo-2.3, MiniMax-Hailuo-02, T2V-01-Director, T2V-01 |
| Image-to-video | MiniMax-Hailuo-2.3, MiniMax-Hailuo-2.3-Fast, MiniMax-Hailuo-02, I2V-01-Director, I2V-01-live, I2V-01 |
What limits does Fast share with 2.3?
On the image-to-video page the duration table gives Hailuo 2.3 and 2.3 Fast the same rows: 6 seconds at 768P (default) or 1080P, and 10 seconds at 768P. Default duration is 6 seconds. Hailuo 02 additionally lists 512P. The page does not state a price or a speed figure for Fast, so any claim about it being cheaper or quicker is not from MiniMax's reference.
The first frame rules are the ones from the same page: JPG, JPEG, PNG or WebP, under 20 MB, short edge above 300 pixels, aspect ratio 2:5 to 5:2.
What do you use on Sume instead?
Sume's video generation docs take a first frame through frame_images with frame_type: first_frame, and the router lists minimax-h3 and minimax-h3-max as supporting image-to-video plus an end frame, 5 to 15 seconds, with native stereo audio. A minimal request uses the same shape as any model:
Set model to minimax-h3, pass one frame_images entry with a public HTTPS image, pick resolution from what GET /v1/videos/models reports, and send an Idempotency-Key so a retry does not bill twice.
What stays unknown
- Whether Fast differs from 2.3 in output quality: MiniMax's reference does not say.
- Whether Fast has a lower price: this page does not list one.
- Whether H3 produces the same motion from the same still: it is a different generation, so rerun your test frames.
- Whether Sume will list a Hailuo 2.3 id: its docs list none today.
Sources
Related posts
More in Models
- MiniMax Hailuo 2.3 vs H3 API: model ids, durations, prompt limits
MiniMax's API still lists MiniMax-Hailuo-2.3 (6 or 10 s, 2,000-character prompt) beside MiniMax-H3 (4-15 s, 7,000). Which one Sume lists, and why it matters.
- Hailuo camera commands: the 15 bracket options like [Push in]
MiniMax's Hailuo API reads 15 camera commands in square brackets, such as [Truck left] or [Push in]. The full list, and what to do on Sume.
- Hailuo first frame image limits: 20 MB, 300 px, 2:5 to 5:2
MiniMax's Hailuo image-to-video wants JPG, PNG or WebP under 20 MB with a short edge over 300 px; H3 frame images list 256-5760 px and 30 MB. Compare both.
- MiniMax video API error 1026 and 1027: sensitive content
MiniMax returns 1026 when the input is flagged and 1027 when the generated video is. What 1002, 1008 and 2013 mean, and how Sume reports failures.
Written by Sume