manyframeJimeng VGFM T2V L20.
A dedicated legacy text-led video identifier. Give the L20 route a clear scene brief and keep the five-second task separate from other Jimeng request keys.
Available models and settings are shown in your studio. Review the quote before generating.
The route.
The right inputs.
Start from the creative job, then choose the model and settings that fit it.
Model ID jimeng_vgfm_t2v_l20
- Duration
- 5 seconds for the declared L20 route
- Seed
- 0 default; supplied seed is forwarded
- Composition
- Aspect ratio is derived from the requested size
- Frame count
- Derived from duration
- Input boundary
- The Responses path accepts text only and rejects image parts
Capabilities describe the Manyframe connector. Your studio shows the currently enabled route, supported settings and maximum charge.
Keep the scene in the written direction.
Describe the subject, the environment and one action that fits a five-second clip. If the destination has a specific composition, choose a supported size before approving the task.
Jimeng's connector can handle several upstream request-key transformations, but those do not become additional declared model IDs. This page covers the exact L20 text identifier. Do not transfer image or newer-duration claims from a different request key into the brief.
Shape a text-led scene ↗
Good questions. Clear answers.
Still curious? Browse all FAQs
Can I send image parts through the Responses route?
No. The Jimeng Responses decoder accepts text input only and rejects image parts.
Do newer Jimeng duration settings apply to L20?
No. The declared L20 request uses five seconds. Other converted request keys are separate behavior, not a capability promise for this identifier.
Make the five-second brief concrete.
Check the exact route, supported size and quote.