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.

For image-to-video on the Hailuo models, MiniMax lists JPG, JPEG, PNG or WebP under 20 MB, a short edge above 300 pixels and an aspect ratio between 2:5 and 5:2 (read 2026-10-02). The H3 guide lists different numbers: first and last frame images of 256 to 5,760 pixels, up to 30 MB per image. If one pipeline feeds both, validate to the stricter value of each limit.
How do the two sets of limits compare?
Both pages are MiniMax's own. The Hailuo numbers come from the image-to-video reference, which also accepts public URLs or Base64 data URLs. The H3 numbers come from the video generation guide, which allows 0 to 2 frame images.
| Limit | Hailuo image-to-video | MiniMax H3 frame images |
|---|---|---|
| Formats | JPG, JPEG, PNG, WebP | Not listed on the guide |
| File size | Under 20 MB | 30 MB per image |
| Dimensions | Short edge above 300 px | 256 to 5,760 px |
| Aspect ratio | 2:5 to 5:2 | Common ratios or adaptive output |
| Count | One first frame | 0 to 2 (first and last) |
Where does Sume differ?
Sume's video generation docs take frame images by URL in frame_images, each with a frame_type of first_frame or last_frame, and troubleshooting asks for images that are public HTTPS and in a supported format. There is no Base64 field in that table, so inline data URLs from a MiniMax integration need to be uploaded first.
The docs do not publish per-model pixel or megabyte limits for minimax-h3, so validate against MiniMax's numbers as the stricter source and expect a validation failure if an input is out of range. If you send both frame_images and input_references, frame_images wins and the call is treated as image-to-video.
What should you check before submitting?
- Convert HEIC or other phone formats to JPG or PNG first.
- Confirm the short edge is above 300 px and the ratio sits inside 2:5 to 5:2; very tall banners fail.
- Keep files well under 20 MB so one limit covers both models.
- Host the file at a public HTTPS URL that does not expire mid-render.
What the pages do not say
MiniMax's image-to-video reference covers the Hailuo models and does not state whether H3 shares its aspect-ratio rule, and the H3 guide does not list accepted formats. Sume lists minimax-h3 for first-and-last-frame use but does not restate either vendor's image limits. Treat each unlisted rule as unknown and test one image before a batch.
Sources
Related posts
More in Models
- 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.
- MiniMax video callback_url: echo the challenge in 3 seconds
MiniMax checks a callback_url by POSTing a challenge you must echo within 3 seconds. Sume skips the handshake and signs each terminal webhook instead.
- MiniMax video tasks are queryable for 7 days: poll every 10 s
MiniMax says video tasks can be queried for the last 7 days and suggests polling every 10 seconds. What that means for your pipeline, and Sume's 30 s example.
- Where are the lyrics in a Lyria 3.5 result? Gemini vs Sume
Google returns Lyria 3.5 lyrics and song structure as text beside the audio. Sume puts model-reported lyrics or a section map in result.lyrics when present.
Written by Sume