Boutique hotel room reel: three 8 s MiniMax H3 clips at 768p
Three 8-second 9:16 room clips on MiniMax H3 at 768p cost $0.60 each on Sume, $1.80 for the set. Plain H3 against H3 Max, and the reference limits.

Three 8-second vertical room clips on MiniMax H3 at 768p cost $0.60 each, $1.80 for the set, on Sume. 768p is the native size of H3 on Sume. Spend more only if the reel will run on a large screen: H3 Max at 768p is $0.80 per clip, and at 1080p $1.60.
What a room reel needs
Guests judge a room by the bed, the window and the light. The usual reel has one clip per space: bedroom, bathroom, terrace. Use real photos of each as image references so the room is the actual room. MiniMax's H3 post, read 2026-10-08, describes image, video and audio references in one model, and Sume's catalog lists up to 9 images per job, 5 to 15 seconds, 480p or 768p, and 9:16 among seven aspect ratios.
H3 sound is native stereo with no toggle on Sume, so expect ambience in the output. Mute it in your editor if the property has its own music.
Cost table
Sume bills the provider list price times 1.25, rounded up to the cent per job.
| Model | Resolution | Per second | 8 s job | Three clips |
|---|---|---|---|---|
| minimax-h3 | 480p | $0.06 | $0.50 | $1.50 |
| minimax-h3 | 768p | $0.07 | $0.60 | $1.80 |
| minimax-h3-max | 768p | $0.10 | $0.80 | $2.40 |
| minimax-h3-max | 1080p | $0.20 | $1.60 | $4.80 |
Staying honest about the room
A reel that shows more than the room has is a booking complaint.
- Send photos of the real room; do not prompt for a view the window does not have.
- Past 5 reference images, plain H3 adds a per-image charge on Sume, so use up to 5 photos per clip.
- Keep camera moves simple: a slow push-in or a pan from bed to window.
- Check the generated furniture count against the room before publishing.
Submit it and read the result
Sume reserves the billable amount from your workspace USD balance when the job is submitted, then returns a polling_url. Run the three rooms as three jobs, each with its own photos. Poll the job until status is completed or failed; a completed job lists its files in unsigned_urls.
import os
import time
import requests
H = {"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}
body = {
"model": "minimax-h3",
"prompt": "A slow push-in on the scene, soft natural light",
"duration": 8,
"resolution": "768p",
"aspect_ratio": "9:16",
}
r = requests.post("https://api.sume.com/v1/videos", headers=H, json=body, timeout=60)
r.raise_for_status()
url = r.json()["polling_url"]
while True:
job = requests.get(url, headers=H, timeout=30).json()
if job["status"] in ("completed", "failed"):
break
time.sleep(10)
print(job["status"], job.get("unsigned_urls"))What I did not verify
I did not test how H3 renders mirrors or glass. The Sume catalog lists 480p and 768p for H3, and 480p, 768p and 1080p for H3 Max; the 2K and 4K tiers MiniMax's page describes for H3 are not offered as separate Sume resolutions.
Sources
Related posts
More in Use cases
- Brewery label that comes alive: Kling 3 first and last frame, 84 cents
Kling 3 on Sume takes a first and last frame, so a 6-second label animation that ends on the finished label costs $0.84 silent, $1.26 with sound.
- Charity appeal video, 20 s on Wan 3.0: $1.25, $2.50 or $5.00
A 20-second appeal on Wan 3.0 costs $1.25 at 480p, $2.50 at 720p and $5.00 at 1080p on Sume. Which tier to publish and how many photos you can send.
- Clean studio look from a phone photo for a listing: edit cost by model
Turn a cluttered phone photo into a clean listing shot with an image-to-image call. 200 edited photos cost $1.93 to $20.00 on Sume depending on the model row.
- Crowdfunding page video: Wan 3.0 hero plus Kling 3 loops, priced
Prototype photos only work as references on Wan 3.0, so it gets the hero. Kling 3 does 1080p feature loops at $0.14 a second. Whole page set: $4.40.
Written by Sume