Pet moment
Image to video · 10s · 768pThe cat in the photo stretches, yawns, and settles back into the sunny spot on the sofa. Camera at cat height. Soft purring.
Start image
AI video model · by MiniMax
MiniMax H3 AI Video Generator
The lowest-priced video engine on GoCrazyAI. Turn a photo into a smooth, expressive clip with stereo sound, or guide it with up to nine reference images, and iterate as often as you like.
- Lowest price
- Native stereo audio
- Image to video
- Up to 9 reference images
- Last-frame control
- Up to 15 seconds
- Video extend
Price
4 credits
for 5s at 480p (≈ $0.76)
Median render
108s
across 112 finished clips, last 60 days
Quality index
76/100
#8 of 9 engines, blind benchmark
Live figures from the GoCrazy Video Index, refreshed hourly. Dollar amounts use the best plan rate.
Made with MiniMax H3
Every clip below was generated on GoCrazyAI with the exact prompt and settings shown. Hover or tap to play.
Anime still to motion
Image to video · 10s · 768pThe anime character in the image turns toward the wind, hair and scarf flowing, cherry petals drifting past. Keep the anime style. Wind chimes.
Start image
Nature loop
Image to video · 10s · 768pLeaves sway and a small stream flows over rocks; sunlight flickers through the trees. Camera fixed. Water and birdsong.
Start image
MiniMax H3 is the newest video model from MiniMax, the team behind the Hailuo engines. It starts from a photo: your image becomes the first frame and the output keeps its shape, so a vertical photo gives a vertical clip without any cropping.
Its appeal is cost. A short clip costs a handful of credits, which makes H3 the engine for exploring ideas: try ten motions for a product shot, pick the best one, then re-render it on a premium engine if you need higher resolution.
It is not a draft-only model, though. Motion is smooth and expressive, faces hold up well, and every clip comes with native stereo audio. Attach up to nine reference images, three clips and three audio files to keep a subject consistent.
What MiniMax H3 can do
- Image to video
- Your photo is the first frame. The clip keeps the photo’s aspect ratio.
- Reference to video
- Up to 9 images, 3 clips and 3 audio files. Reference clips are trimmed to 15 seconds in total and audio to 15 seconds each.
- Last-frame control
- Add an end image and the clip lands on it.
- Native audio
- Stereo sound is always generated with the picture.
- Extend
- Continue a clip by 5, 10 or 15 seconds at 480p or 768p.
- Clip length
- 5, 10, 15 seconds
- Resolution
- 480p, 768p
MiniMax H3 price per clip
Credits, from the live price list. You always see the exact cost before you generate.
| Resolution | 5s | 10s | 15s |
|---|---|---|---|
| 480p | 4≈ $0.76 | 7≈ $1.33 | 10≈ $1.90 |
| 768p | 7≈ $1.33 | 13≈ $2.47 | 19≈ $3.61 |
Dollar figures use the best plan rate. See plans and credit packs.
Where MiniMax H3 shines
Cheap enough to explore
The lowest credit cost per clip in the studio, so you can try several motions and camera moves before committing to a premium render.
Expressive motion
Natural gestures and facial expression, with smooth camera moves that rarely jitter at the edges of the frame.
Stereo sound included
Every clip ships with a stereo track generated alongside the video, at no extra cost.
Keeps your framing
The output follows your photo’s aspect ratio, so vertical, square and wide photos all animate without cropping.
Nine references
Guide a character or product with up to nine images, plus reference clips and audio, for consistency across a series.
Quick turnaround
Short clips at 480p or 768p render fast, so feedback loops stay tight while you refine a prompt.
How to use MiniMax H3
- 1
Open AI Video Pro
MiniMax H3 is preselected when you arrive from this page. Add a start image, or leave it empty to generate from text, or attach references.
- 2
Write the shot
Describe the subject, the action, the camera and the light in one or two sentences. Start from a tested prompt below.
- 3
Pick length and generate
Choose duration, resolution and aspect ratio. The credit cost updates as you change them, and failed renders are refunded.
MiniMax H3 prompts to start with
Six prompts written for how MiniMax H3 reads direction. The full guide has 30 more, sorted by use case.
Product idea test
The product in the photo rotates slowly on a turntable while soft light sweeps across it. Locked-off camera, clean background. Low ambient hum.
Cheap to try several variations before a premium render.
Portrait comes alive
The woman in the photo looks up from her book, smiles and tucks her hair behind her ear. Gentle window light, camera still. Quiet cafe sounds.
Vertical street scene
People walk past a busy street corner at dusk; the man in the photo checks his watch and crosses the road. Handheld feel. City ambience.
Use a vertical photo for a vertical clip.
Nature loop
Leaves sway and a small stream flows over rocks; sunlight flickers through the trees. Camera fixed. Water and birdsong.
Pet moment
The cat in the photo stretches, yawns, and settles back into the sunny spot on the sofa. Camera at cat height. Soft purring.
Anime still to motion
The anime character in the image turns toward the wind, hair and scarf flowing, cherry petals drifting past. Keep the anime style. Wind chimes.
MiniMax H3 vs the other engines
Same studio, same credits. Switch engines per clip; here is when another one fits better.
- Seedance 2.5: when you need more than 15 seconds or want to start from text with no image.
- Wan 3.0 Prime: the step up to 1080p and 30 seconds, with ten reference images.
- Seedance 2.0: when the final delivery has to be 1080p or 4K.
- HappyHorse 1.1: for multi-subject scenes built from several reference images.
| Engine | Longest clip | Audio | Credits / 5s | Median render |
|---|---|---|---|---|
| Seedance 2.5 | 30s | Native | 30 | 385s |
| Seedance 2.5 Turbo | 30s | Native | 16 | — |
| Wan 3.0 Prime | 30s | Native | 10 | 88s |
| MiniMax H3 | 15s | Native | 4 | 108s |
| Seedance 2.0 Mini | 15s | Native | 10 | 227s |
| Seedance 2.0 Fast | 15s | Native | 15 | 213s |
| Seedance 2.0 | 15s | Native | 20 | 242s |
| Wan 2.7 | 15s | Native | 17 | 114s |
| HappyHorse 1.1 | 15s | Native | 19 | 110s |
Credits for a 5-second clip at 720p (or the engine's nearest tier). Full method on the Video Index.
Frequently asked questions
What is MiniMax H3?
MiniMax H3 is a video model from MiniMax, the company behind the Hailuo engines. It turns a photo into a clip of up to 15 seconds with native stereo audio, and can follow up to nine reference images.
Is MiniMax H3 the cheapest video engine on GoCrazyAI?
It is the lowest-priced engine per clip in AI Video Pro. The price table on this page comes from the live price list, and the generator shows the exact cost before you confirm.
What resolution does MiniMax H3 output?
480p or 768p. The clip follows the aspect ratio of your photo. For 1080p use Wan 3.0 Prime or Seedance 2.5 Turbo.
Can MiniMax H3 make video from text only?
No. It needs a start image or at least one reference. For prompt-only video use Seedance 2.5 or HappyHorse 1.1.
How long can a MiniMax H3 clip be?
5, 10 or 15 seconds per generation. Extend a clip afterwards by another 5 to 15 seconds.
Does MiniMax H3 have sound?
Yes, every clip includes native stereo audio generated with the picture.
Can I use my own photos?
Yes. Uploads are checked automatically before generation. GoCrazyAI is for adults (18+) only, and anyone in your photos must be an adult who has agreed to appear.
Make your first MiniMax H3 clip
Opens AI Video Pro with MiniMax H3 selected. Paste a prompt from this page and adjust the length before you generate.
Last updated: