Omni Flash rejects 768p: the resolution strings each video model takes
Omni takes 360p, 720p, 1080p or 4K; H3 takes 768p, not 720p. One table of the accepted resolution strings per model and the price range of each on Sume.

Sending 768p to Gemini Omni Flash 1.1 fails, because Omni prices only 360p, 720p, 1080p and 4K, while MiniMax H3 prices 480p, 768p, 2K and 4K. The price check says so in words: Cannot price resolution: expected 360p, 720p, 1080p, or 4K.
The table below lists the accepted strings for the four models in this lane and the per-second range of each, from $0.0375 on Omni 360p up to the top row (Sume price tool, read 2026-10-09).
Accepted resolutions and per-second range
The match is case-aware for the K tiers: the pricing code accepts 4K and the lowercase 4k as an alias on Omni, and 2K/4K or lowercase on H3.
| Model | Accepted resolutions | Duration | Cheapest per second | Dearest per second |
|---|---|---|---|---|
| Gemini Omni Flash 1.1 | 360p, 720p, 1080p, 4K | 3-10 s | $0.0375 | $0.375 |
| MiniMax H3 | 480p, 768p, 2K, 4K | 5-15 s | $0.0625 | $0.20 |
| MiniMax H3 Max | 480p, 768p, 1080p | 5-15 s | $0.0625 | $0.20 |
| H3 Max Recast | 768p, 1080p | 5-30 s source | $0.375 | $0.5625 |
Where the near-misses come from
The cheapest fix for a near-miss is to read the allowed list from the catalog before you build the request.
- 720p sent to MiniMax: use 768p. MiniMax 768p is native and first-class, not a 720p.
- 1080p sent to MiniMax H3: H3 has no 1080p row, so use H3 Max.
- 768p sent to Omni: use 720p (or 1080p).
- 480p sent to Recast: Recast takes 768p or 1080p and defaults to 768p.
Read the list instead of guessing
GET /v1/videos/models returns supported_resolutions and supported_durations for every video model. Compare your request to those lists in your own code and you avoid the round trip.
import os, requests
r = requests.get("https://api.sume.com/v1/videos/models",
headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}, timeout=30)
r.raise_for_status()
for model in r.json()["data"]:
if model["id"] in ("gemini-omni-flash-1.1", "minimax-h3", "minimax-h3-max"):
print(model["id"], model["supported_resolutions"])The same 8 seconds, four ways
The clearest way to see the ladders is to price the same 8 seconds at the nearest resolutions. Omni has no 768p and MiniMax has no 720p, so 720p and 768p are different rows on different models.
| Model | Resolution | Rate per second | 8 s total |
|---|---|---|---|
| Omni | 720p | $0.125 | $1.00 |
| MiniMax H3 | 768p | $0.075 | $0.60 |
| MiniMax H3 Max | 768p | $0.10 | $0.80 |
| Omni | 1080p | $0.1875 | $1.50 |
| MiniMax H3 Max | 1080p | $0.20 | $1.60 |
Limits and gotchas
A 2K or 4K request to MiniMax H3 is priced, but it is an upscale of native 768p, not a native render. Omni 4K is its own tier at its own price.
Sources
Related posts
More in Models
- Python clip planner for a Short: split any length into model jobs
A short Python function splits a Short's length into equal whole-second jobs within each Sume video model's min and max, then prices them with Timeline.
- Qwen Image Max at $0.09375 vs Nano Banana 2.1 at $0.10: 100 images
On Sume, Qwen Image Max costs $0.09375 per image and Nano Banana 2.1 costs $0.10 at 1K. 100 images are $9.38 or $10.00. Ratios, resolution tiers and the pick.
- Recast 480p fails: resolution must be 768p or 1080p, priced by length
H3 Max Recast refuses 480p with 'resolution must be 768p or 1080p'. A 10-second source costs $3.75 at 768p and $5.625 at 1080p on Sume. The 5 to 30 s table.
- Seedream 5 Lite 4:5 portrait: render 1920x2400, resize to 1080x1350
Seedream 5 Lite lists 4:5 on Sume at $0.04375 per image. A 1920x2400 render shrinks to Instagram's 1080x1350 at an exact 0.5625 scale. 100 images cost $4.38.
Written by Sume