Long AI video clips on Sume: which rows pass 15 seconds
Only two Sume rows pass 15 seconds in one job: Seedance 2.5 (4-30 s) and Wan 3.0 (2-30 s). All other rows stop at 15 s or less. Read 2026-10-13.
If the clip must be longer than 15 seconds in one job, use seedance-2.5 (4 to 30 seconds) or wan-3.0 (2 to 30 seconds). Every other standard row stops at 15 seconds or less: Kling 3 and the H3 rows at 15, Seedance 2 and its Fast and Mini variants at 15, Grok at 15, Omni at 10. sume/auto is limited to 3-10 seconds, so a long request must pin an id.
Row ceilings
Genjutsu and Recast also run up to 30 seconds, but they edit a source clip rather than generate from a prompt.
| Row | Longest clip |
|---|---|
seedance-2.5 | 30 s |
wan-3.0 | 30 s |
kling-3, H3, H3 Max, seedance-2*, Grok | 15 s |
gemini-omni-flash-1.1 | 10 s |
| Veo rows | 8 s |
Beyond 30 seconds
Chain clips: end one at the frame where the next begins, using an end-frame row. See the loops post for frame inputs, and plan for continuity drift between jobs.
Sources
Related posts
More in Models
- MAI-Voice-2.1 emotion control vs Sonic's six emotions on Sume
Microsoft lists granular emotion control for MAI-Voice-2.1. On Sume, Sonic 3 rows take six named emotions in generation_config.emotion; Eleven rows have none.
- Can Meta Muse Wonder generate images? It returns text; use /v1/images
Muse Wonder is listed on OpenRouter as a chat-completions model with text responses. It is not in Sume's image catalog; pick an image row instead.
- Decision-1 GlobalStandard or DataZoneStandard for a Sume gate?
Microsoft lists both deployment types for Decision-1. DataZoneStandard keeps processing in a data zone; GlobalStandard may vary more in latency. Benchmark.
- Mistral Large 4 function calling: validate arguments against Sume
Mistral Large 4 (Public Preview) lists function calling and structured outputs. Validate its arguments with Sume's tools_schema before any paid call.
Written by Sume