Seedance 2.5
#11 in Text-to-BildFal · v2.5 · 3× · tolest 12. Aug. 2026
39
Momentum
Seedance 2.5 is a multimodal video generation model developed by ByteDance (Dreamina/BytePlus) and provided via fal through API. It generates native single-take videos up to 30 seconds long with synchronized audio, optionally from text, an image, or up to 50 multimodal references (image, video, audio, style). On fal, the model is available through three endpoints (Text-to-Video, Image-to-Video, Reference-to-Video), with output in 480p or 720p in six aspect ratios. Billing is Token-based; separate LoRA Fine-Tuning for Seedance 2.5 itself is not documented on fal (LoRA training is offered there for other models such as MiniMax H3).
Momentum-Verloop
14.05.12.08.
Features
| Fine-Tuning | No dedicated LoRA fine-tuning for Seedance 2.5 documented on fal; LoRA training on fal only available for other models (e.g. MiniMax H3) |
| Max Resolution | 720p (480p or 720p selectable) |
| Max Video Length | 30 seconds native, single continuous take (selectable 4-30 sec) |
| Platform | Available via fal (Playground, Sandbox, API, MCP server, CLI, fal Agent) |
| Price | $0.0214 per 1,000 tokens; approx. $0.4730/sec at 720p, approx. $0.2205/sec at 480p |
| Release Date | Model launch July 31, 2026; fal API live from August 2026 |