Omni vs Seedance 2.5 vs Wan 3.0 for an 8-second sound-on ad
Price, length range and audio handling for an 8-second ad on Gemini Omni Flash 1.1, Seedance 2.5 and Wan 3.0 through Sume, with a pick-by-job rule.
For one 8-second ad, Omni Flash costs $1.00 at 720p and $1.50 at 1080p on Sume, Wan 3.0 costs $1.00 and $2.00, and Seedance 2.5 costs $4.6224 and $11.3724. Pick by what the clip has to do, not by the lowest number.
Price and limits side by side
| Model | 8 s at 720p | 8 s at 1080p | Duration | Audio |
|---|---|---|---|---|
| gemini-omni-flash-1.1 | $1.00 | $1.50 | 3-10 s | Native, always on |
| wan-3.0 | $1.00 | $2.00 | 2-30 s | Audio toggle (Alibaba page) |
| seedance-2.5 | $4.6224 | $11.3724 | 4-30 s | Check catalog generate_audio |
Choosing
Alibaba's page for wan3.0-video lists 480P, 720P and 1080P, 2 to 30 seconds, MP4 with audio and an audio toggle (read 2026-10-08). Omni has no audio off switch: its audio is always on, so you omit generate_audio for it.
- Need sound that is guaranteed present, up to 10 s: Omni.
- Need sound optional or a length over 10 s: Wan 3.0 or Seedance 2.5.
- Need audio or video references fed into the model: the Seedance 2.x models and Wan 3.0 accept audio references; Omni does not.
Per second
Divide the 8-second figures by eight and the 720p rates are $0.125 per second for Omni and for Wan 3.0, and $0.5778 per second for Seedance 2.5. That is a factor of about 4.6, so a pick of Seedance 2.5 should come from what it does for the clip, not from habit.
Swap test
All three use the same request. Run the same prompt with each model id at 720p, compare the cuts, then spend the 1080p money on the winner.
Sources
Related posts
More in Comparisons
- One 1920x1080, 24 fps, 20 s master for Amazon, LinkedIn and Threads
A single 20-second 16:9 MP4 at 24 fps sits inside the listed rules for Amazon Sponsored Brands video, LinkedIn video ads and the Threads API. The Sume settings.
- One 30 s shot or three 10 s shots: same price on Seedance 2.5
On Sume, three 10 s Seedance 2.5 shots cost the same as one 30 s shot within 1 cent, plus $0.10 to join them. So cut by story, not by price.
- OpenAI realtime mini ($10/$20 per M audio tokens) vs a narration job
OpenAI lists gpt-realtime mini at $10 in and $20 out per million audio tokens, and tts-1 at $15 per million characters. What each suits, and where Sume fits.
- OpusClip Pro MCP connector vs the Sume hosted MCP: what differs
OpusClip lists a Video Editing API, Scheduler API and MCP connector on Pro. Sume has a hosted MCP with trim and inspect tools. What each gives an agent.
Written by Sume