MiniMax H3 768p vs Gemini Omni Flash 720p: an 8-second clip
On Sume, an 8-second clip is $0.60 on minimax-h3 at 768p and $1.00 on gemini-omni-flash-1.1 at 720p, with $0.30 at Omni's 360p. Price and constraints.
An 8-second clip costs $0.60 on Sume's minimax-h3 at 768p ($0.075 x 8) and $1.00 on gemini-omni-flash-1.1 at 720p ($0.125 x 8). That is a $0.40 gap. Omni's 360p tier is $0.30 for the same 8 seconds, and wan-3.0 at 720p is $1.00. The gap buys different things, described below the table.
Same clip, five rows
All rates are Sume catalog prices. The note column gives the constraint to check in GET /v1/videos/models.
| Model and tier | Per second | 8-second clip | Constraint |
|---|---|---|---|
| minimax-h3, 480p | $0.0625 | $0.50 | 5 to 15 s |
| minimax-h3, 768p | $0.075 | $0.60 | 5 to 15 s |
| gemini-omni-flash-1.1, 360p | $0.0375 | $0.30 | 3 to 10 s, 16:9 or 9:16, audio always on |
| gemini-omni-flash-1.1, 720p | $0.125 | $1.00 | 3 to 10 s, 16:9 or 9:16, audio always on |
| wan-3.0, 720p | $0.125 | $1.00 | 2 to 30 s |
Reading the constraints
Omni is limited to 16:9 and 9:16 and always generates native audio, so the $1.00 includes a soundtrack whether you want one or not. If you need a square clip, Omni cannot produce it on Sume. minimax-h3 has a 5-second minimum, so a 3-second clip is out of range there.
The Omni 360p row is the cheapest listed second on Sume at $0.0375, but 360p is a small frame; treat it as a preview tier.
Choosing
For a silent social loop where 768p is enough, minimax-h3 at $0.60 is the cheaper end. For a clip that needs sound generated with the picture and is 3 to 10 seconds, Omni at $1.00 is the same price as wan-3.0 at 720p and adds audio. Which model looks better for your prompt is something to test at the cheap tiers first.
Sources
Related posts
More in Comparisons
- MiniMax H3 Max vs Wan 3.0 at 1080p: 5, 10 and 15 seconds on Sume
On Sume, MiniMax H3 Max at 1080p costs $0.20 a second and Wan 3.0 costs $0.25. For 5, 10 and 15 seconds: $1.00/$2.00/$3.00 vs $1.25/$2.50/$3.75.
- MiniMax H3 Max 768p: 15 s is $1.50 on Sume, $1.20 on Runway Dev
Runway Dev lists h3_max at 5 credits per second at 480p and 8 at 768p. Sume lists minimax-h3-max at $0.0625 and $0.10, plus a 1080p row at $0.20.
- MiniMax Hailuo 2.3 at 56 cents vs H3 at 80 cents, 10 seconds of 768P
MiniMax's page lists Hailuo 2.3 under Legacy Models: $0.56 for a 10-second 768P clip, $0.32 for Fast. H3 is $0.80. Sume lists H3 only.
- Nano Banana 2.1 and GPT Image 2.5 both list $30 per million tokens
Google and OpenAI both list $30 per million image output tokens. Google's per-image prices imply 1,120, 1,680 and about 3,767 tokens for 1K, 2K and 4K.
Written by Sume